Professional Documents
Culture Documents
Getting Started With Cloud Computing:: A LITA Guide
Getting Started With Cloud Computing:: A LITA Guide
Getting Started
with Cloud
Computing:
A LITA Guide
Edward M. Corrado
Heather Lea Moulaison
Editors
Guide #16
All rights reserved. Reproduction of this book, in whole or in part, without written
permission of the publisher, is prohibited.
The paper used in this publication meets the minimum requirements of American National
Standard for Information Sciences—Permanence of Paper for Printed Library Materials,
ANSI Z39.48-1992.
Getting started with cloud computing : a LITA guide / edited by Edward M. Corrado,
Heather Lea Moulaison ; with a foreword by Roy Tennant.
p. cm. — (LITA guide ; #16)
Includes bibliographical references and index.
ISBN 978-1-55570-749-1 (alk. paper)
1. Cloud computing. I. Corrado, Edward M., 1971- II. Moulaison, Heather Lea,
1974- III. Library and Information Technology Association (U.S.)
QA76.585.G48 2011
004.6782—dc23
2011017261
51046_ch00_FM_pi-xviii.qxd 5/19/11 12:45 PM Page iii
Contents
List of Illustrations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
Foreword
Roy Tennant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii
Contents
vi Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
Implementation Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
Social Networking within SharePoint . . . . . . . . . . . . . . . . . . . . . . . . . . 135
SharePoint Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
Selection Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
Assessment and Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
List of Illustrations
FIGURES
Figure 2.1 Amazon Web Services Bill . . . . . . . . . . . . . . . . . . . . . . . . . 17
Figure 2.2 Amazon Web Services’ Modify Database
Instance Screen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Figure 2.3 Hypothetical SFX (Link Resolver) Systems
Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Figure 2.4 Hypothetical Redmine Systems Architecture
in the Cloud . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Figure 9.1 IDS Gift Deselection Manager . . . . . . . . . . . . . . . . . . . . . . 99
Figure 9.2 VuFind Recommender Module—Authors Related
to Your Search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 ix
Figure 9.3 LibX—Library Catalog Configuration . . . . . . . . . . . . . . . . 100
Figure 9.4 Peer Review Indicator Prototype . . . . . . . . . . . . . . . . . . . . 101
Figure 9.5 Miami University Library Cite This Button . . . . . . . . . . . 102
Figure 9.6 Results from Libraries Worldwide . . . . . . . . . . . . . . . . . . . 103
Figure 9.7 iBookshelf Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
Figure 9.8 CampusBooks Application . . . . . . . . . . . . . . . . . . . . . . . . . 105
Figure 9.9 Pic2Shop Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
Figure 10.1 Amazon Web Services Sign In . . . . . . . . . . . . . . . . . . . . . . 113
Figure 10.2 The AWS Management Console Dashboard . . . . . . . . . . . 113
Figure 10.3 Generating an RSA Key Pair . . . . . . . . . . . . . . . . . . . . . . . 114
Figure 10.4 Searching for the DRC AMI . . . . . . . . . . . . . . . . . . . . . . . . 115
Figure 10.5 Instance Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Figure 10.6 Review Options and Launch . . . . . . . . . . . . . . . . . . . . . . . 116
Figure 10.7 Public DNS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
51046_ch00_FM_pi-xviii.qxd 5/19/11 12:45 PM Page x
TABLES
Table 4.1 Factors and Issues with Cloud Platforms . . . . . . . . . . . . . . 41
Table 4.2 Example Distribution of Services on Cloud Platforms . . . 42
Table 6.1 Recommended Content Management Systems . . . . . . . . . 65
Table 6.2 Recommended Library, Archive, and Repository Systems 66
Table 6.3 Recommended Public Knowledge Project Software . . . . . 67
x Table 13.1 Comparison of Costs for OCLC versus
Amazon S3 for 5 TB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
51046_ch00_FM_pi-xviii.qxd 5/19/11 12:45 PM Page xi
Foreword
F rom one perspective, the history of computing can be viewed as a cyclical gy-
ration of centralization versus decentralization of computing power. The first
machines were too expensive for any organizations or individuals except govern-
ment agencies and eventually very large businesses. Centralization was broken by
the advent of the personal computer, which put computing power in the hands of
individuals for the first time. The era of the minicomputer consolidated decen-
tralization, as each individual organization installed “dumb” terminals lashed to a
minicomputer, including the library where I worked in 1983.
Then came the Internet, which provided a way to interconnect all of these
nodes. This enabled another wave of centralization, as we increasingly accessed
servers so far away and anonymous that we never knew where they actually sat.
To some degree, cloud computing represents taking centralization to its logical
conclusion. With cloud computing, the very idea of “clients” and “servers” fades xi
into the background. You have a phone, or a pad device, or a laptop, or a desktop,
and it is able to do many wondrous things because it is seamlessly integrated over
the network with farms of servers that operate transparently and that automati-
cally grow with demand.
It is indeed a brave new world we live in, and this book demonstrates that li-
brarians are not only aware of these changes but are already using them to more ef-
fectively and efficiently fulfill their needs and the needs of their users. In that way
this book is both a lesson and an inspiration. It is an inspiration as it demonstrates
the resiliency of librarians in acquiring new skills and applying new technologies
to serve the needs of their users ever better. It is a lesson because in these pages are
both foundationally instructive pieces as well as case studies of exactly how to
bring these new technologies to bear in your own organizations.
This is no different from the era in which I became a librarian, when the Inter-
net was beginning to transform our world. Both instructional explication and case
studies were needed, to illustrate both what could be accomplished with this new
tool as well as exactly how librarians were using it. Therefore, you now have in
your hands the chapter and verse on what cloud computing is and how to use it ef-
fectively in your library.
51046_ch00_FM_pi-xviii.qxd 5/19/11 12:45 PM Page xii
For me the most powerful aspect of cloud computing is that it enables libraries
to stop dealing with technical issues that have nothing to do with their day-to-day
mission and services. Cloud computing offers the ability to simply requisition
technological resources that can expand as required to meet expanding needs
while requiring little or nothing from the library providing the service. Addition-
ally, using cloud computing services it is possible to get a full-featured website
going in minutes—I’ve done this multiple times myself using Amazon’s EC2 ser-
vice and I’m still astonished. It quite literally has never been easier to get a website
going than it is today.
So within these pages rests the future of libraries. We must get out of the busi-
ness of buying, configuring, installing, and maintaining servers unless we ab-
solutely must do so. Otherwise, we should be simply and easily requisitioning
services from the cloud, and immediately and effectively putting them into ser-
vice. After all, this is exactly what everyone else is doing.
Roy Tennant
Senior Program Officer
OCLC Research
xii
51046_ch00_FM_pi-xviii.qxd 5/19/11 12:45 PM Page xiii
Preface
Preface
Lastly, Part III: Case Studies (Chapters 13–20) gives insights into projects that
libraries throughout the United States have undertaken. Each chapter features sub-
sections giving a background to the library and to the problem. Chapter sections
describe the cloud-based tools considered, how the selected solution was imple-
mented, and an evaluation of the end product. In Chapter 13, Edward Iglesias
begins by describing how his library saves archival quality images in the cloud. In
Chapter 14, Karen A. Reiman-Sendi, Kenneth J. Varnum, and Albert A. Bertram
describe their experience integrating the cloud-based LibGuides with their locally
hosted website by taking advantage of Springshare’s XML output feature. In the
following case study in Chapter 15, Caitlin A. Bagley details the use of Dropbox
for file sharing between and among librarians and for the remote access of library
instruction documents by embedded librarians at her institution. In Chapters 16
and 17, respectively, Anne Leonard and Robin Elizabeth Miller both describe ex-
periences using free software from Google: Leonard describes benefits and uses of
Google Calendar, and Miller describes innovative uses of Google Forms for in-
struction. In Chapter 18, Leland R. Deeds, Cindy Kissel-Ito, and Ann Knox de-
scribe their experience using Ning both in the library and for distance education at
their college. In Chapter 19, Ann Whitney Gleason describes using DimDim for
web conferencing. Although the DimDim software is no longer on the market,
Gleason’s experiences with a cloud-based web conferencing tool can still inform li-
brarians’ thought processes as they consider other web conferencing tools for li-
brary services. Lastly, in Chapter 20, Jennifer Ditkoff and Kara Young describe
their uses of VoiceThread for instruction. xv
Although the chapters in this book are written in such a manner that they can
stand on their own individually, Getting Started with Cloud Computing is designed
to be a complete volume that is possible to read from cover to cover. The chapters
and indeed the parts of the book build on each other; readers are encouraged to
start from the beginning and read their way sequentially through the book’s chap-
ters. However, if readers are trying to implement a cloud-based solution similar
to one of the case studies, they should feel free to jump in at the point that makes
the most sense, as readers would do with any technical handbook covering new
technologies.
The editors hope this book paves the way for easier implementation of cloud
computing and helps readers think about cloud computing in libraries as they
never have before.
51046_ch00_FM_pi-xviii.qxd 5/19/11 12:45 PM Page xvi
51046_ch00_FM_pi-xviii.qxd 5/19/11 12:45 PM Page xvii
Acknowledgments
xvii
51046_ch00_FM_pi-xviii.qxd 5/19/11 12:45 PM Page xviii
51046_ch01_p001-012.qxd 5/19/11 12:47 PM Page 1
Part I
GENERAL CONCERNS
51046_ch01_p001-012.qxd 5/19/11 12:47 PM Page 2
51046_ch01_p001-012.qxd 5/19/11 12:47 PM Page 3
Chapter 1
Perspectives on Cloud
Computing in Libraries
Heather Lea Moulaison
and Edward M. Corrado
Introduction
During the first decade of the current millennium, the term “cloud computing”
has become almost ubiquitous in certain circles. Gmail is in the cloud. Facebook is
in the cloud. Remotely hosted integrated library systems (ILSs) are also in the
cloud. The term “cloud” is on everyone’s lips, yet, depending on the community,
the term may well mean different things both in theory and in practice. While busi-
nesses may look to the cloud for low-cost solutions to data storage and computing 3
power problems, libraries may be interested in targeted solutions for sharing data
with other libraries and providing low-cost services to users. Many types of com-
puting problems can benefit from the use of cloud computing solutions, but those
solutions may involve a different set of considerations for libraries due in part to
the nature of their mission and activities.
Different models for deploying the cloud have made cloud computing an inter-
esting proposition for businesses as well as for nonprofits, and those involved with
libraries and cultural heritage institutions are right to explore their options. This book
is designed to help librarians and administrators think about cloud computing as a
powerful technology, to consider some tools that are available for use in libraries, and
to learn about the experiences of libraries already using these technologies.
Cloud computing is designed to give users flexibility for dealing with their
computing needs. Because of the inherent elasticity that is part of the cloud com-
puting model, cloud computing is especially useful in industries where there is a
sharp rise and decline in activity. Holiday shopping activities are classic examples
of a smart way that online merchants make use of cloud computing. Companies
offering online purchasing may not have enough computing power locally to sup-
port web sales in December, but it does not make economic sense for them to pur-
chase additional servers and hire additional staff year-round for the single month
per year that the computing power is necessary. Instead, these companies are
better served to host some aspect of their operations in the cloud and to let their
providers bill them for the use incurred on a monthly basis. If they were so
inclined, librarians could certainly arrange for a similar computing power to be
available to them to support their diverse computing needs.
Cloud computing can also be beneficial when expertise is lacking or when
future computing needs are unsure. A large percentage (43.3 percent) of respon-
dents from the Zenoss open source systems management community indicated
that flexibility is the main reason they choose to use virtualization software
(Zenoss, 2010: 6). Librarians, especially at smaller libraries, may see the same
advantages in cloud computing. Ambitious librarians wanting to start new services
may also appreciate using resources in the cloud, because, with cloud computing,
there is “[t]he elimination of an up-front commitment by cloud users, thereby
allowing companies to start small and increase hardware resources only when
4 there is an increase in their needs” (Armbrust et al., 2010: 51).
Some librarians are excited about cloud computing because it offers an avenue
for better cooperation. Using a data as a service (DaaS) model, for example, it
would be possible for libraries to better cooperate to maintain bibliographic and
authority records. As Karen A. Coombs describes in Chapter 9, one way this can
be done is by taking advantage of application programming interfaces (APIs).
The cloud also allows libraries to cooperate with other libraries and library
software vendors to secure rights to article metadata from publishers and database
providers. With a few possible exceptions for large or well-funded libraries, nego-
tiating access to locally store this metadata and then make it available to end users
in a meaningful manner would likely be cost prohibitive. However, as Marshall
Breeding elaborates on in Chapter 7, some library software vendors are now able
to provide discovery services, utilizing metadata provided by many different con-
tent providers, to libraries that offer a more complete, single search experience for
library patrons.
On-demand computing power is not the only reason to consider turning to
the cloud. Other reasons to use cloud computing include hardware savings (33.3
percent), administrative ease of use (8.7 percent), and labor cost savings (6.0 per-
cent) (Zenoss, 2010: 6). If librarians can save on buying and maintaining servers,
for example, they can save their parent institutions money. Depending on the ser-
51046_ch01_p001-012.qxd 5/19/11 12:47 PM Page 5
vice, they can also rely on the software provider to carry out all updates, thereby
allowing their own employees to take care of other obligations in lieu of monitoring
version changes and upgrades. Furthermore, quite a few of the cloud-based soft-
ware platforms are available for free, meaning that librarians can make use of ser-
vices without incurring any cost for their libraries. The promise of saving money
and easing the administrative workload is enticing in the for-profit sector, but is
especially interesting for libraries and cultural heritage institutions that may be
increasingly forced to find ways to decrease the bottom line while continuing to
offer the same or even better services.
Working in organizations that may not have recourse to legal counsel and in ones
that will not have the kind of large accounts that American Express has, librari-
ans may find themselves on their own for thinking through some of the issues of
cloud computing. This can be somewhat daunting given that both costs and oblig-
ations need to be understood and considered (Chaput and Ringwood, 2010: 241).
Indeed, this book is intended to get librarians and administrators thinking about
using cloud computing services, but it is understood that each library will need
to assess the cloud landscape for itself before making decisions to rely in any way
on cloud computing.
Some library administrators may be interested in cloud computing because of
the perceived cost savings in staff. It is true that with many, if not most, cloud com-
puting solutions there is no need to purchase or maintain servers, and there are
associated cost savings in that area. However, the added fees paid to the provider
may or may not outweigh the costs of a non-cloud-based solution. Also, while you
may not need someone to administer a server, there is most likely still a need for
someone to administer the software application itself. Migrating to a cloud-based
application may turn out to be cost neutral but at the same time still have the ben-
efit of allowing staff to concentrate more on library-specific issues.
At present, many libraries in the United States have free or low-cost accounts
for software that is hosted in the cloud or pay to host data that is not sensitive. For
example, several chapters in the Technologies section and the Case Studies section
address free and low-cost options for software in the cloud. Chapters in the Gen-
6 eral Concerns section and the Technologies section describe the use of hosted
solutions either through a library vendor or through a standard account available
to any user. This volume does not feature chapters describing instances where
libraries are hosting sensitive data in the cloud quite simply because none have
come to the attention of the editors.
bility and interoperability, Chaput and Ringwood (2010: 252) remind us, “Plan-
ning for contingency is paramount for outsourcing arrangements and cloud
arrangements are no different. It is important to identify who owns the data and
ensure that both parties agree.” For further thoughts on this question, see Carl
Grant’s chapter (Chapter 5) for the point of view of a vendor/librarian.
Risk
There is a certain amount of risk any time sensitive data is recorded and entrusted
to any party for any reason. While it may be easier to trust those who are known, it
51046_ch01_p001-012.qxd 5/19/11 12:47 PM Page 9
may not be possible to ensure that data is truly any more secure within the
metaphorical walls of the library than beyond them. Using cloud services “is nei-
ther inherently insecure nor secure” (Chaput and Ringwood, 2010: 241). As with
the Ma.gnolia and JournalSpace.com examples earlier, there is a risk both to hous-
ing library data in the cloud (the risk the users of the websites took) and to using
internal staffing (the risk the owners of the websites took).
There are other risks with cloud computing besides employee sabotage or
error. Librarians will want to consider the possibility of the cloud computing
provider’s going out of business. Is that provider the only place the library’s data is
housed? If so, will the library be able to get the data back, and, in the case that it
can, will it be in a timely manner? Librarians contemplating moving to a cloud-
based service will need to determine what level of risk they are willing to take and
what those risks are compared to hosting the application in-house. The answers
will differ depending on the organization and on the cloud provider.
Data Ownership
Data in the cloud brings many concerns such as the aforementioned privacy and
security implications. Another concern libraries need to be aware of is data own-
ership. Questions regarding data ownership need to be addressed before embark-
ing on a cloud computing project—especially one that places data and applications
critical to day-to-day library operations such as an integrated library system (ILS)
into the cloud. Who owns the data placed in the cloud, and what rights do the
library and provider have to the data? After the contract expires, how and in what 9
format will the data be returned to the library, or how will the library retrieve the
data? Libraries should make sure that any contract they sign with a cloud provider
includes information about access to the data on an ongoing basis and at the end of
the contract. The process and time frame for receiving data in an agreed-upon
format should be delineated (Trappler, 2010).
Evaluation
While this book does not address the concept of evaluation of cloud services
directly, it acknowledges that evaluation is an essential component of any new ser-
vice’s implementation in libraries. The editors of this guide recommend that eval-
uation of cloud computing be carried out the way any other evaluation is carried
out at a reader’s library. Although the evaluation component offered in the Case
Studies section tends to be informal, it is a first step toward getting libraries to
quantify their cloud computing experiences and to share that information outside
the walls of their institution.
Conclusion
Cloud computing has generated a great deal of buzz in technology circles in gen-
eral and within the library community in particular for good reason. The advan-
51046_ch01_p001-012.qxd 5/19/11 12:47 PM Page 10
tages of cloud computing include flexibility, ease of use, cost savings on hardware,
and possible time savings for staff that would allow technology staff to concen-
trate on tasks more closely related to the library mission than the maintenance of
servers.
Cloud computing, however, is not a panacea that will cure all library technol-
ogy ills. While libraries have ebbs and flows in computing needs, they are not at
the same scale as an Internet retailer during the holiday season. There are various
concerns that librarians need to consider when moving services, especially mis-
sion-critical services, to the cloud. If the network connection goes down or is
unbearably slow, staff and patrons may not be able to access necessary applica-
tions, data, or other content entrusted to the cloud computing service. There are
various data security and privacy concerns that are specific to libraries and that
need to be addressed by library institutions on an individual basis. Also, librarians
need to consider when first implementing a cloud-based solution how they will
get their data out of the cloud should they need or want to change providers.
Libraries are entrusted with content that must be safeguarded and protected, and
they must provide services to help users access that content. When content and
services are outsourced to third-party cloud computing providers, librarians may
feel that they have lost a degree of control that would be maintained had they
hosted their content and services themselves.
Questions of control are mitigated when there is a sense of trust in the profes-
sionalism of the cloud computing provider. While there are certainly privacy and
10 security risks with cloud computing, it is not necessarily any more risky than run-
ning one’s own servers. In fact the cloud may be more secure, depending on the
availability and skill level of local technology staff. Hackers can compromise a local
server just as easily as a cloud-based one if the proper security steps have not been
taken. Librarians need to take a critical look at cloud providers to see how much
trust they should be afforded and should act accordingly.
For better or for worse, the cloud allows librarians to experiment with tech-
nology like never before and with little or no up-front costs. For example, instead
of having to purchase servers to house a DSpace repository, libraries can outsource
their computing needs to Amazon as John Davison describes in Chapter 10.
Economies of scale can be exploited, partner libraries can be brought on board,
and projects can be expanded, scaled back, and changed with little to no inconve-
nience to an institution’s librarians.
Librarians and library administrators, especially those interested in library
technology, have reason to be excited about cloud computing and the possibili-
ties it offers. As with any other service or product, however, librarians need to eval-
uate cloud-based solutions to see if they offer a good return on investment and if
they further the library’s mission. The chapters that follow are designed to guide
the reader in that quest for understanding and for finding inspiration.
51046_ch01_p001-012.qxd 5/19/11 12:47 PM Page 11
References
Anderson, Kent. 2010. “Is the Cloud Too Weak to Support What Paper Can?” The Schol-
arly Kitchen (blog), December 8. http://scholarlykitchen.sspnet.org/2010/12/08/
is-the-cloud-too-weak-to-support-what-paper-can/.
Armbrust, Michael, Armando Fox, Rean Griffith, Anthony D. Joseph, Randy Katz, Andy
Konwinski, et al. 2010. “A View of Cloud Computing.” Communications of the ACM
53, no. 4: 50–58. doi:10.1145/1721654.1721672.
Chaput, Shawn R., and Katarina Ringwood. 2010. “Cloud Compliance: A Framework for
Using Cloud Computing in a Regulated World.” In Cloud Computing: Principles, Sys-
tems and Applications, edited by N. Antonopoulos and L. Gillam, 241–255. London:
Springer-Verlag London Limited. doi:10.1007/978-1-84996-241-4_14.
Dawson, Sophie, and Andrew Hilton. 2010. “Privacy Changes Could Affect Cloud Pro-
jects.” CIO, December 20. http://www.cio.com.au/article/370772/privacy_changes
_could_affect_cloud_projects/.
Halff, Larry. 2009. “Whither Ma.gnolia?” Citizen Garden 11. Video. 25:04, uploaded Feb-
ruary 13. http://vimeo.com/3205188.
Humphries, Matthew. 2009. “Disgruntled Employee Kills JournalSpace with Data Wipe.”
Geek.com (blog), January 5. http://www.geek.com/articles/news/disgruntled-employee-
kills-journalspace-with-data-wipe-2009015/.
Iowa State University. 2009. “Iowa State Researcher Says Internal Security Breaches Pose
a Bigger Threat than Hackers.” Iowa State University News Releases, July 31. http://
www.news.iastate.edu/news/2009/jul/security.
Kavur, Jennifer. 2010. “Don’t Use the PATRIOT Act as an Excuse.” IT World Canada,
July 5. http://www.itworldcanada.com/news/dont-use-the-patriot-act-as-an-excuse/ 11
141033.
Obama, Barack. 2010. “Remarks by the President on the Ongoing Oil Spill Response.”
Transcript, and MP4 Video, 7:20, uploaded May 14, The White House.
http://www.whitehouse.gov/blog/2010/05/14/relentless-efforts-stop-leak-and-contain-
damage.
PrivacySense.net. 2010. “The Canadian Equivalent to the PATRIOT Act.” Privacy-
Sense.net. Accessed December 23. http://www.privacysense.net/canadian-equivalent-
patriot-act/.
Trappler, Thomas J. 2010. “If It’s in the Cloud, Get It on Paper: Cloud Com-
puting Contract Issues.” EDUCAUSE Quarterly Magazine 33, no. 2. http://
www.edu-cause.edu/EDUCAUSE+Quarterly/EDUCAUSEQuarterlyMagazineVolum/
IfItsintheCloudGetItonPaperClo/206532.
Wispinski, Jennifer. 2006. “The USA PATRIOT Act and Canada’s Anti-Terrorism Act:
Key Differences in Legislative Approach.” Parliamentary Information and Research
Service (PIRS) of the Library of Parliament, March 31. http://www2.parl.gc.ca/
Content/LOP/ResearchPublications/prb0583-e.pdf.
Zenoss. 2010. 2010 Virtualization and Cloud Computing Survey. Annapolis, MD: Zenoss.
51046_ch01_p001-012.qxd 5/19/11 12:47 PM Page 12
51046_ch02_p013-028.qxd 5/19/11 12:47 PM Page 13
Chapter 2
Introduction
Cloud computing is the new “it” thing in the information technology community.
Everyone from Microsoft to Google to Amazon purport to have some type of cloud
offering. Microsoft recently released a series of advertisements in which people,
faced with difficult tasks, go “to the cloud” to accomplish what they need. In one
advertisement a couple stuck in an airport uses a laptop to remotely watch TV
shows from their home computer. In another, a mother uses Windows to merge
multiple family portraits into one, creating a more idyllic, singular portrait.
Google has quietly offered a number of cloud-based solutions to users for quite 13
some time. These include Google Docs and its even more popular YouTube ser-
vice. But, Google’s offerings do not stop there. Its popular Google App Engine
allows developers to create and host applications on Google’s data centers, mini-
mizing development costs and allowing a company to get their application out the
door more quickly. Their latest offering, Google Storage for Developers, provides
“a RESTful [Representational State Transfer] service for storing and accessing
your data on Google’s infrastructure” (Google, 2010), minimizing the costs asso-
ciated with storing large quantities of data and creating redundancies to prevent
data loss.
The cloud provider that often eclipses them all is Amazon. With its myriad
services, Amazon has one of the most robust offerings of any of the vendors out
there. Using Amazon’s Simple Storage Solution (S3), subscribers can store as
much data as they need. They can also provision a server in just a few clicks using
Amazon’s Elastic Compute Cloud (EC2). Or hook into a MySQL database using
its Relational Database Service (RDS).
On-Demand Self-Service
When an institution makes a decision to implement a system, it generally works
with a vendor to make the system operational. For example, a library might decide
51046_ch02_p013-028.qxd 5/19/11 12:47 PM Page 15
to purchase an integrated library system (ILS); one of the first likely steps would
be to purchase a server for the ILS. After consulting with the ILS vendor to deter-
mine the requirements for the server, the library will call on a number of hardware
vendors, provide them with the server specifications, and get price quotes. Once
the vendor and price have been determined, the vendor’s sales representative will
place an order for the server and the server will be shipped to the library. When it’s
received, the library will need to rack the server, connect it to its network, and
install the operating and supporting software.
When implementing the same system in the cloud, the process for purchasing
a server is radically different. Although the institution would still need to deter-
mine the server requirements from the ILS vendor, the rest of the process would
change. For cloud-based server providers, rates are usually posted on their web-
sites, thereby eliminating the need to work with a sales representative to get a price
quote. In the case of some providers, like Amazon Web Services, there is even an
online calculator to help the client project its monthly bill (Amazon, 2010c).
Ordering a server would no longer be an obstacle either. Instead, cloud
providers allow their clients to create an account, attach it to a credit card, and
begin purchasing services immediately. Rackspace, which allows clients the ability
to provision servers on demand, provides a simple to use form that walks clients
through the process of creating an account. Once it is created, access to its cloud-
based servers is available instantly. Users simply need to find the server in the list of
available templates, provision it, and then log into the server to begin using it. Con-
trast this to the hours it can take to set up a locally hosted system that requires 15
network drops, racking a server, and installing the server’s base operating system.
This new process can save both network and system administrators a consider-
able amount of time.
wireless device” (Rainie, 2004). Whereas six years later a similar report indicates,
“59% of American adults now go online wirelessly using either a laptop or cell
phone” (Smith, 2010). In the earlier report there is only one mention of using
mobile phones for access to the Internet, but in the later there are 164. The way we
use thin devices and the way we access the Internet has radically changed and led
to the evolution of cloud computing.
So what does it mean that cloud-based services are available through broad
network access? The NIST asserts that “capabilities are available over the net-
work and accessed through standard mechanisms that promote use by heteroge-
neous thin or thick client platforms” (Mell and Grance, 2010). This means that
users can use cloud-based services through any type of Internet-ready device,
unlike in the past when users were required to download an application in order to
work with the computing service they had purchased.
Being able to access computing resources over the Internet is just the tip of
the iceberg. In the past, methods for accessing computing services were usually
offered only through the service provider. For instance, to upload a picture to a
particular provider’s website, one was required to download and install an appli-
cation created and distributed by the provider itself. Today, APIs have made it
possible for third parties to create their own applications for use with the
providers’ services. The Android Market illustrates this brilliantly; the market con-
tains dozens of web applications available for Flickr users, none of which was cre-
ated by Flickr. Most, if not all, cloud services provide some type of API so that
16 users can access the service through whatever means they want.
The most prominent example of a service available through broad network
access is Google Docs. In 2007, Google launched Google Docs, which allowed
users to create and edit documents via the web. Traditional word processors and
spreadsheet editors required users to install software locally. Google Docs, on the
other hand, linked users to existing Google accounts, creating a space for them to
produce and collect documents using the web-based software, eliminating the need
to install it locally. Over the years, Google Docs has become more full featured,
adding new upgrades like charts for spreadsheets and collaboration tools for docu-
ments. Because the software itself is available through any web browser, Google
can make upgrades transparently and seamlessly for its users. In 2010, Microsoft
released its Office Web Apps, a recent competitor to Google Docs. This trend indi-
cates that the future of word processing may very well occur over the network.
Metered Service
There are countless books, blog posts, and articles out there comparing cloud com-
puting to electricity. The most prominent is Nicholas Carr’s (2008) book The Big
Switch. In it he discusses the rise of electricity as we know and use it today. He dis-
cusses, in fascinating detail, how Edison originally saw electricity as something that
would be created and used strictly on-site in a business or home. It was his clerk,
51046_ch02_p013-028.qxd 5/19/11 12:47 PM Page 17
Samuel Insull, who saw the bigger picture. He created a system of generators and a
delivery network, thus making electricity a public utility. Carr goes on to argue that
the same thing is happening with computing. Instead of services being located on-
site, we are farming computing out to a select few providers (Carr, 2008). While
many argue that the analogy has been taken too far (Urquhart, 2009), one place
where the comparison becomes readily apparent is in the metering of services.
Outsourcing is not a new concept for technology. People have been using vir-
tual private servers and hosting providers for years. The pricing structures for
these services are either by the month or year or through some other contract the
institution sets with the hosting provider. Generally, the users of these services are
limited in the amount of computing resources they can consume. For example, a
customer may be allowed to use only 350 GB of bandwidth each month and 100
GB of storage. If they go over these limits the hosting provider may deny users
access to the website or turn the service off altogether. These old pricing struc-
tures are similar to Edison’s view of electricity: you use only what you have, and
you can’t have any more without a significant investment.
The cloud model is different, though; you pay according to your use. This means
that if users use 3 TB of storage the first month but only 1 TB in the second month,
the bill will be adjusted accordingly. In previous payment models, users would be
charged for 3 TB of space regardless of whether or not they were using that space.
Figure 2.1 provides an example of a bill from Amazon Web Services. In it,
users are charged for hours a device is running, I/O (input/output) requests to the
device, storage on the device, storing an image of the device, and any GETs made 17
to the device. Similar to a public utility, subscribers are charged based on which
resources are used and how much they are being used. While the metered pay-
ment model is not a new one to the business world, it is new to the computing
industry and is changing the way institutions provide computing services.
Elasticity
Setting up a new system can often be a guessing game. Common questions we
might ask vendors or the institution are:
• How much storage is needed?
• How much RAM is needed?
• What kind of redundancy should be implemented?
• Will the system grow?
• How process intensive is the system?
The purpose of asking these questions is to better understand the type of infra-
structure needed to implement a system.
Let’s say that a large library system plans on implementing a new search tool.
They hope to first release it to particular graduate school libraries; if the project is a
success, they will release it to the entire campus. The user base at the beginning of
the system’s implementation may be dramatically different when compared to the
end of its implementation; the infrastructure developed must take this into account.
At the beginning of the project, the team is faced with one of two decisions:
1. Create a system that will work with the pilot project. Perhaps the team
can reuse an existing server. Later, if the project is a success the team
will need to determine how to expand the infrastructure while main-
taining any changes or keeping data they may have added to the system.
18 2. Create a system that will work for the eventual population. This may
include purchasing new servers, backup tools, or anything else the infra-
structure might need. Essentially the team is planning on the success of
the project and spending money accordingly.
Neither of these options is ideal. Many institutions now deploy virtual servers that
allow for expansion and contraction, but sometimes institutions are still faced with
purchasing expensive hardware to ensure that a system can be expanded when
needed.
The cloud, however, offers the ability to expand without the worry of pur-
chasing hardware or other components to improve the system’s infrastructure.
This means that an institution can quickly expand from a system that works for a
thousand users to a system that works for tens of thousands.
An example of a cloud-based service that allows for quick and easy expansion
is Amazon Web Services’ Relational Data Service (RDS). With RDS, a user can
create a MySQL database and later choose to expand its space. Figure 2.2 shows
the modify database instance screen available through the Amazon Web Services
Management Console. Here are a number of options available to users. One allows
users to change the database from a small instance, which according to Amazon’s
website is “1.7 GB memory, 1 ECU (1 virtual core with 1 ECU), 64-bit platform,
Moderate I/O Capacity,” to a four times large instance that has “68 GB of memory,
51046_ch02_p013-028.qxd 5/19/11 12:47 PM Page 19
26 ECUs (8 virtual cores with 3.25 ECUs each), 64-bit platform, High I/O Capac-
ity” (Amazon, 2010a). Still another option allows users to specify the amount of
storage needed for the data. This can be as small as 5 GB or as large as 1 TB.
These features and others make Amazon’s RDS easily scalable simply by clicking
a few buttons or making a few API calls.
Resource Pooling
Traditionally when an institution purchases infrastructure for a system, it must
purchase according to the system’s needs during peak loads. Consider a library
purchasing a new discovery interface. Administrators know that during midterms 19
and finals the system will be under an intense load, so they purchase a server
based on the load they expect during those four time periods. During the rest of
the year, they will only utilize a quarter of the server’s capacity. In the end, the
institution is wasting money by not taking advantage of the infrastructure’s true
potential.
If the same institution’s administrators choose to use the cloud, they could
provision a server based on their normal use case. They could then utilize APIs or
other services to scale the system up or down based on the load to the server. The
rest of the time the computing power they are not using is released back into the
general pool to be used by other institutions.
Perhaps another institution’s administrators use the cloud to provision a
number of high-capacity servers to process large quantities of data. This work is
something that they need to do once a year, and purchasing servers just for this
work would be a fiscal waste. By using the cloud, the institution provisions the
servers, processes the data, and releases the servers back into the general pool of
resources.
The idea of sharing computing resources with the larger pool is often referred
to as multitenancy: “a single instance of software [that] runs on a server, serving
multiple client organizations (tenants)” (Wikipedia, 2010). The best example of
51046_ch02_p013-028.qxd 5/19/11 12:47 PM Page 20
will run the end-user software. In the case of SFX this consists of the program-
ming language Perl, the Apache web server, the open source database MySQL,
and the operating system Ubuntu. Once these pieces are in place and function-
ing, the software itself can be installed on top of the platform and infrastructure.
IaaS doesn’t just stop with servers though. Amazon and other providers also
offer a number of storage and network solutions. For example, Google recently
announced its Google Storage for Developers, combining “the performance and
scalability of Google’s cloud with advanced security and sharing capabilities”
(Google, 2010), and Rackspace, a major competitor to Amazon, provides users
with on-demand IP addresses that can easily be associated or disassociated with
any of the “servers” customers provision. These infrastructure services are dra-
matically changing the way users are creating and implementing technology.
A Note
It is important to remember that many do not believe that these three service
models are the only possible service models. In his book Cloud Computing with
Windows Azure Platform, Roger Jennings (2009) outlines other service models,
including files as a service, data as a service, communication as a service, and mon-
itoring as a service. While it may be possible to provide any type of computing
resource as a service, it is also possible to broaden these ideas and fit them into
one of the three NIST-defined service models.
• Public cloud
• Community cloud
• Private cloud
• Hybrid cloud
The following section will move through each of the deployment models, provid-
ing examples and discussing issues and controversies associated with the different
models as well as how one might deploy a system within that model.
Public Cloud
A public cloud is a cloud service that is available for use by the general public.
Users of this type of cloud need to simply sign up for an account and the service is
made available to them instantly. Users would make payments to the cloud
provider based on their usage of the different services. Some examples of public
cloud providers are Google, Amazon, Microsoft, and Rackspace.
If an institution was interested in setting up a system in Amazon’s cloud it
might first search through the available Amazon Machine Images, or AMIs, to see
if an image of that system already exists. If it does, the institution can easily launch
the system with a few simple clicks. If it does not, the institution can choose an
51046_ch02_p013-028.qxd 5/19/11 12:47 PM Page 24
AMI that contains the proper supporting platform or it can choose to build that
platform itself. Once it has that platform in place, it can install the software and
deploy its new system.
There are, of course, a number of concerns that institutions have with deploy-
ing systems into a public cloud; chief among these is security. What information
can we trust cloud providers with? Legally, what information are we allowed to
trust cloud providers with? It is important for any institution that is looking to
move to the cloud, or even to outsource a system, to properly assess the risk
involved in giving up some portion of control over that system.
There are a number of services that different cloud providers offer their users
in order to help make their systems more secure. Amazon offers a Virtual Private
Cloud that allows users “to connect their existing infrastructure to a set of isolated
AWS compute [sic] resources via a Virtual Private Network (VPN) connection,
and to extend their existing management capabilities such as security services,
firewalls, and intrusion detection systems to include their AWS resources”
(Amazon, 2010b).
In its Cloud Computing Risk Assessment report, the European Network and
Information Security Agency (ENISA, 2009) outlines a number of security bene-
fits for the cloud:
• Security and the benefits of scale
• Security as a market differentiator
24 • Standardized interfaces for managed security services
• Rapid, smart scaling of services
• Audit and evidence gathering
• More timely, effective, and efficient updates and defaults
• Benefits of resource concentration
Deploying applications to a public cloud may bring up many questions and con-
cerns for an institution. It is important that, before implementing a system in the
cloud, institutions thoroughly evaluate it, the cloud provider, and the data they
plan on storing in that system.
Community Cloud
Community cloud is cloud infrastructure that “is shared by several organizations
and supports a specific community that has shared concerns (e.g., mission, secu-
rity requirements, policy, and compliance considerations). It may be managed by
the organizations or a third party and may exist on premise or off premise” (Mell
and Grance, 2010). The most commonplace community clouds are governmental
clouds; the U.S. government, the European Union, and a number of other gov-
ernmental bodies around the world use the cloud in order to share resources
among their different agencies (Di Maio, 2010). In higher education, community
clouds are already beginning to appear. OSHEAN, the Rhode Island Research
51046_ch02_p013-028.qxd 5/19/11 12:47 PM Page 25
and Education Network, is in the process of building and expanding its Cumulus
Cloud Computing (C3), which allows OSHEAN members to “annually subscribe
to a dedicated Resource Pool” (OSHEAN, 2010).
Community clouds offer many advantages over public clouds. The payment
structure for a community cloud might differ greatly from that of the public cloud.
Institutions may pay an annual fee based on FTE or may pay based on their annual
usage; all of this would depend on how the community chooses to set up payment
structures. Additionally, security concerns might be reduced when using a com-
munity cloud based on prior relationships the institution may have with the com-
munity offering the cloud service. And, within a community cloud, institutions
may feel they have a greater voice in how the cloud itself is operated and secured.
Educational institutions and libraries, in particular, are perfectly poised to pro-
vide community clouds. These institutions have been working together, creating
vast networks of cooperation among one another. The idea of sharing computing
resources is not a new one either, and creating a community cloud is simply the
next step in sharing with one another.
Private Cloud
Private cloud is by far the most controversial of all the deployment models. Private
cloud is defined as an individual institution operating its own cloud. The cloud
can be operated either by the institution or by a third party, and it can be hosted at
the institution or off-site (Mell and Grance, 2010).
Many argue that a private cloud is not the cloud. In the blog post, “When Is a 25
Cloud Not a Cloud?” Phil Wainewright (2009) points out that cloud computing
comes from the notion of the Internet and the countless diagrams using clouds to
represent the Internet. When a private cloud is launched at an institution, it is
essentially captive or cordoned off from the rest of the Internet. Wainewright
(2009) argues that this is the primary reason that it is not the cloud. It doesn’t
need the Internet to run; just its local network will suffice.
Conversely, private cloud allows institutions to reap the benefits of the cloud
(metering, scalability, on-demand self-service, etc.) but without the risks (security,
lack of control, and so on). Established computing vendors like Oracle, IBM,
EMC2, and VMWare all offer some type of private cloud-based service in con-
junction with their traditional computing offerings, helping to legitimize the pri-
vate cloud as a true cloud deployment model.
Hybrid Cloud
Hybrid cloud allows institutions to deploy an application or system using more
than one type of deployment model. Figure 2.4 is an example of how an institu-
tion might deploy Redmine, an issue/software management system, in the cloud.
Perhaps in this case the institution is interested in keeping control of the data main-
tained in the MySQL database but isn’t as interested in hosting the files associated
51046_ch02_p013-028.qxd 5/19/11 12:47 PM Page 26
with the application. The institution decides to use VMWare’s vCloud service to
host the database and Amazon EC2 for the remainder of the system’s architecture.
Hybrid cloud, though, can go beyond the normal decoupling IT shops are
used to implementing. “Eucalyptus is an open-source software platform that
implements IaaS-style cloud computing using the existing Linux-based infra-
structure found in the modern data center” (Eucalyptus, 2010). When users
implement Eucalyptus in their data center they receive an experience similar to
that of an Amazon Web Services user. Eucalyptus also integrates with EC2, allow-
ing users to employ cloud bursting, which is
26 the practice of “bursting” into a cloud when capacity has been reached in the cor-
porate cloud/data center. The business case for cloud bursting primarily revolves
around seasonal or event-based peaks of traffic that push infrastructure over its
capacity but are not consistent enough to justify the cost of investing in additional
hardware that would otherwise sit idle. (MacVittie, 2009)
When Amazon released its Virtual Private Cloud, or VPC, data centers everywhere
stood up and cheered. This new offering provided data center operators with the
ability to place greater control on Amazon’s cloud offerings, essentially moving
the public cloud under the umbrella of the institution’s data center.
Hybrid cloud allows users of the cloud to integrate the different deployment
models, public, community, and private, into one seamless cloud. Users are then
able to benefit from all of the things each of these deployment models has to offer.
Conclusion
The NIST definition of cloud computing offers a sound foundation of under-
standing for anyone interested in learning more about the cloud. The five charac-
teristics of the cloud help users recognize what attributes cloud services possess.
The four deployment models provide users with an understanding of where they
can set up their systems. Finally, the three service models supply a framework of
how providers deliver these services to their users.
51046_ch02_p013-028.qxd 5/19/11 12:47 PM Page 27
References
Amazon. 2010a. “Amazon Relational Database Service (Amazon RDS).” Amazon Web
Services. Accessed November 7. http://aws.amazon.com/rds/.
———. 2010b. “Amazon Virtual Private Cloud.” Amazon Web Services. Accessed Novem-
ber 7. http://aws.amazon.com/vpc/.
———. 2010c. “Simple Monthly Calculator.” 2010. Amazon Web Services. Accessed
November 7. http://calculator.s3.amazonaws.com/calc5.html.
Anderson, Tim. 2010. “Microsoft’s Office Web Apps—Google Killing Not Included.” The
Register (London, UK), April 27. http://www.theregister.co.uk/2010/04/27/office
_2010_web_apps_review/.
Carr, Nicholas G. 2008. The Big Switch: Rewiring the World, from Edison to Google. New
York: W.W. Norton & Co.
Di Maio, Andrea. 2010. “Community Clouds from Governments to Banks Will Challenge
Vendors.” Gartner Blog Network (blog), April 27. http://blogs.gartner.com/
andrea_dimaio/2010/04/27/community-clouds-from-governments-to-banks/.
ENISA (European Network and Information Security Agency). 2009. “Cloud Comput-
ing Risk Assessment.” ENISA. November 20. http://www.enisa.europa.eu/act/ 27
rm/files/deliverables/cloud-computing-risk-assessment.
Eucalyptus. 2010. “Products Overview.” Eucalyptus. Accessed November 7.
http://www.eucalyptus.com/products/overview.
Google. 2010. “Google Storage for Developers—Google Code.” Google Code. Accessed
November 7. http://code.google.com/apis/storage/.
Heroku. 2010. “How It Works.” Heroku. Accessed November 7. http://heroku.com/
how/architecture.
Hoff, Chris. 2009. “On the Draft NIST Working Definition of Cloud Computing. . . .”
Rational Survivability (blog), May 8. http://www.rationalsurvivability.com/blog/
?p=870.
Hurwitz, Judith. 2010. Cloud Computing for Dummies. Hoboken, NJ: Wiley.
Jennings, Roger. 2009. Cloud Computing with the Windows Azure Platform. Indianapolis,
IN: Wiley.
MacVittie, Lori. 2009. “Cloud Balancing, Cloud Bursting, and Intercloud.” F5 DevCentral
(blog), July 9. http://devcentral.f5.com/weblogs/macvittie/archive/2009/07/09/cloud-
balancing-cloud-bursting-and-intercloud.aspx.
———. 2010. “Architectural Multi-tenancy.” F5 DevCentral (blog), May 18. http://
devcentral.f5.com/weblogs/macvittie/archive/2010/05/18/architectural-multi-
tenancy.aspx.
Mather, Tim, Subra Kumaraswamy, and Shahed Latif. 2009. Cloud Security and Privacy.
Beijing: O’Reilly.
51046_ch02_p013-028.qxd 5/19/11 12:47 PM Page 28
Mell, Peter, and Tim Grance. 2010. “Cloud Computing.” National Institute of Standards
and Technology. Last updated August 27. http://csrc.nist.gov/groups/SNS/cloud-
computing/.
OSHEAN. 2010. “Cumulus Cloud Computing (C3) Service, by OSHEAN.” OSHEAN.
Accessed November 7. http://cumulus.oshean.org/.
Rainie, Lee. 2004. “The Rise of Wireless Connectivity and Our Latest Findings.” Pew
Research Center’s Internet & American Life Project. April 13. http://www.pewtrusts
.org/our_work_report_detail.aspx?id=22660&category=54.
Smith, Aaron. 2010. “Mobile Access 2010.” Pew Research Center’s Internet & American
Life Project. July 7. http://pewinternet.org/Reports/2010/Mobile-Access-2010.aspx.
Urquhart, James. 2009. “In Cloud Computing, Data Is Not Electricity.” The Wisdom of
Clouds. CNET News. August 1. http://news.cnet.com/8301-19413_3-10296370-
240.html.
Wainewright, Phil. 2009. “When Is a Cloud Not a Cloud?” ZDNet. August 28.
http://www.zdnet.com/blog/saas/when-is-a-cloud-not-a-cloud/846.
Wikipedia. 2010.”Multitenancy.” Wikipedia. Revised October 13. http://en.wikipedia.org/
wiki/Multitenancy.
28
51046_ch03_p029-036.qxd 5/19/11 12:48 PM Page 29
Chapter 3
Cloud Computing:
Pros and Cons
H. Frank Cervone
Introduction
Cloud-based computing is a hot topic right now because many people believe that
it will dramatically transform the way we use and provide access to technology.
Others, however, are not so sure (Farber, 2008). Nonetheless, while perhaps not as
transformational as the Internet itself, cloud computing will provide an evolution-
ary step in the provision of network-based services by changing the way people
access applications and other services. This emphasis on services is important as it
is the main attraction of moving to cloud computing. While the underlying tech-
nology is interesting to technologists, it is ease of access and provisioning of ser- 29
vices that is the real benefit to organizations.
As is often the case with newer technologies, as many people and organiza-
tions begin to adopt the new technology, there is a lot of confusing information
swirling around, and much of it is based on subjective opinion rather than objec-
tive reasoning or facts. Cloud computing is no exception. Today, much of the
information related to cloud computing approaches the topic from either one of
two perspectives. The first approach takes the position that it is the savior of all the
ills in the information technology world. The other perspective is that cloud com-
puting is a bad idea and will lead to the inevitable failure of an organization if
adopted. While these depictions may be a bit exaggerated, they demonstrate the
two polar opposite approaches to cloud computing being sold today. However,
neither of these perspectives is completely correct.
The research firm IDC predicts that cloud computing will move from a tech-
nology used by “early adopters” to one that is part of everyday information tech-
nology operations by 2012 (Gens, 2008). Therefore, the goal of this chapter is to
present a balanced view of both the pros and cons of using cloud computing so
that librarians and employees in other information organizations can make
informed decisions on how and when to proceed. Not surprisingly, as is the case
with most technologies, the usefulness of cloud computing will depend greatly on
51046_ch03_p029-036.qxd 5/19/11 12:48 PM Page 30
In addition to these points, Heiser and Nicolett (2008) of the Gartner Group
suggest the following security points be considered:
1. External audit and security certification of the cloud-based
provider. Ultimately, an organization is responsible for the security and
integrity of its data, regardless of where it is held. Service providers
must undergo external audits and security certifications for the good of
the library.
2. Storing and processing data in specific jurisdictions. A library may
be subject to laws and regulations that govern where data can (and,
more importantly, cannot) be stored. Librarians will need to work with
their provider to both understand and ensure that their data is being
stored in legal locations.
3. Investigative support. This is critical in certain circumstances. The
cloud provider must be able to ensure that if the data is subpoenaed or
requested for other investigation, it will be able to produce it within a
reasonable time frame.
4. Provider long-term viability. If the cloud services provider being used
34 goes bankrupt or is acquired, librarians must ensure that their data will
be accessible after such an event.
Conclusion
While still new, adoption of cloud computing is gaining steam because most orga-
nizations see that the advantages of controlling all aspects of information technol-
ogy implementations are outweighed by the benefits of outsourced models that
can provide a greater range and diversity of services than could be provided in-
house. In addition to reducing the startup costs for new services, cloud computing
can be more cost effective in the long term as services are usually based on a pay-
as-you-go model, similar to utilities. By reducing costs, cloud computing allows
organizations to use their resources to focus on the truly unique aspects of service
they need to provide to their community.
This is not to say that there are not concerns associated with cloud computing.
Issues related to security, pricing, and availability need to be addressed before any
large-scale deployment of cloud-based services. Vendors need to be carefully eval-
uated to ensure they have the capability and resources to deliver the services
promised. Additionally, organizations need to consider the long-term viability of
their cloud services providers.
51046_ch03_p029-036.qxd 5/19/11 12:48 PM Page 35
Many libraries will find that introduction of cloud-based services will come
on a piecemeal basis as their traditional vendors experiment with this new model.
By introducing cloud-based services one at a time, libraries can assess the appro-
priateness and value of such services. Proceeding in this manner also helps address
concerns related to vendor lock-in. If librarians dislike the service, it is likely that it
can be swapped out with a competing service that is more in tune with the local
context.
Perhaps the most compelling argument for cloud-based computing is that
moving to the cloud computing model will help libraries become more agile and
better able to compete in an uncertain environment by eliminating the overhead of
technology support and allow for a greater focus on service provision and innova-
tion.
References
Brenner, Bill. 2010. “Reset.” CIO Magazine 24, no. 2: 30–36. http://www.cio-digital
.com/ciodigital/20101015.
Farber, Dan. 2008. “Oracle’s Ellison Nails Cloud Computing.” CNET News. September
26. http://news.cnet.com/8301-13953_3-10052188-80.html.
Gens, Frank. 2008. “IT Cloud Services User Survey, Pt. 1: Crossing the Chasm.” IDC
Exchange (blog), September 29. http://blogs.idc.com/ie/?p=205.
Heiser, Jay, and Mark Nicolett. 2008. “Assessing the Security Risks of Cloud Computing.”
Gartner Research. June 3. http://www.gartner.com/DisplayDocument?id=685308.
Kim, Won. 2009. “Cloud Computing: Status and Prognosis.” Journal of Object Technology 35
8, no. 1: 65–72. http://www.jot.fm/issues/issue_2009_01/column4/.
Ryan, Vincent. 2008. “A Place in the Cloud.” CFO Magazine. September 1.
http://www.cfo.com/article.cfm/11954936.
Sheard, Reed. 2010. “Cloud Computing in Education: A Practitioner’s View.” Campus
Technology. Accessed December 28. http://campustechnology.com/Articles/
2010/09/22/Cloud-Computing-in-Education-A-Practitioners-View.aspx?p=1.
Wittman, Art. 2010. “Practical Analysis: Our Maturing View of Cloud Computing.” Infor-
mation Week. Accessed December 28. http://www.informationweek.com/story/
showArticle.jhtml?articleID=224701815.
51046_ch03_p029-036.qxd 5/19/11 12:48 PM Page 36
51046_ch04_p037-046.qxd 5/19/11 12:48 PM Page 37
Chapter 4
Introduction
This chapter examines the role that cloud computing has played in transforming
libraries and information technology organizations. It begins with a history of tech-
nology adoption in libraries and concludes with an exploration of current trends in
cloud computing and how these trends are beginning to transform libraries.
change. In the early 1980s, the personal computer (PC) market was driven by
word processing and other localized productivity applications. Over the past 24
years mobility and network connectivity have played key roles in transforming the
most popular applications to cloud-based versions of traditional productivity
applications. A key example of this shift can be seen in the use of Google Docs as
opposed to traditional productivity application formats. Research (Low, 2010) on
users who have both Google Docs and traditional productivity suites installed on
their machine show that they are more frequently using the Google products but
that time spent in an application may depend on the type of application (e.g.,
Microsoft Excel outweighed Google Spreadsheets in time spent).
During this time information formats also changed. For example, the most com-
monly accessed information services on smartphones are communication based,
not publication based (A. Smith, 2010). These formats include text messaging, e-
mail, Facebook, and Twitter. While there are many publication formats accessed
via mobile environments, they often include point-of-need information such as
weather, transit availability, or location information. Likewise information struc-
tures changed from primarily narrative-based text to a mix of multimedia, struc-
tured metadata, and text-based information. This shift extends from client-based
computing to larger scale data analytics. Geoffrey Moore (2010) asserts that use
analysis now requires a cloud computing orientation: “The IT stack is changing
from being data centric to cloud centric—you cannot do metadata analytics at scale
in a data centric environment.” The shift in computing platform and in informa-
38 tion structure is joined with a shift in appliance use. Cloud-based service providers
such as Skype (http://skype.com) represent an abstracted means of accessing voice
and video communication services, and file synchronization sites such as Dropbox
(http://dropbox.com) and Jungle Disk (http://jungledisk.com) provide cloud-based
disk services that both abstract the service from the hardware and provide a suite of
advanced services that are not easily deployed in a client-hardware environment
(e.g., version control, continuous data protection, automatic replication).
Finally, the cost of technology has driven innovation and change. Not only has
technology continued to follow the pace of advancement suggested by Moore, who
said that circuits on a chip double every 24 months (Intel, 2010), the cost of com-
puting resources has dropped considerably—$200/MB for storage in 1980 com-
pared with $.08/GB for storage in 2010 (I. Smith, 2010). As this drop in price
has occurred, service subscription models have been created that do not rely
wholly on hardware cost to provide computing services. For example, the Amazon
AWS service platform (http://aws.amazon.com) provides computing resources,
disk space, notification services, and monitoring services at by-the-hour pricing
models. As of November 2010, the cost per month for 1 GB of disk space was
$.14 per month. While considerably more expensive than comparable costs for
local hard-drive space, Amazon is providing advanced services such as automatic
redundancy, version history, and guaranteed uptime. The same trend is occurring
51046_ch04_p037-046.qxd 5/19/11 12:48 PM Page 39
SERVICE
TYPE INFLUENTIAL FACTORS IMPORTANT ISSUES
APPROACH SYSTEMS
sets. Second, libraries often lack the large-scale infrastructure that would enable
them to deploy new services without considerable capital investment. By selecting
PaaS and IaaS solutions libraries can gain access to these resources on a subscrip-
tion basis without capital investment. Finally, libraries have to find a balance
between services with very rapid life cycles (e.g., video tutorial sites) and services
with very stable life cycles (e.g., integrated library systems). While it can be diffi-
cult to find a single IT service management (ITSM) approach that equally sup-
ports these services, ITSM approaches that leverage cloud-based solutions with
appropriate SLAs and management policies are good matches with how libraries
42 approach IT. Table 4.2 shows an example of how a library might approach the
distribution of its services based on these factors.
apps, and social reward applications. The impact of this transition is that not only
is our approach to accessing computer resources changing, but the ways in which
these resources are used to create, share, and leverage information are also chang-
ing. While predicting the evolution of the cloud computing movement so early in
its life cycle is ultimately not particularly productive, it is worth thinking about
some large-scale changes that cloud computing may introduce into information-
centric organizations. Three areas in particular stand out as being potentially
affected by cloud computing changes. These areas are personal empowerment, IT
issue redefinition, and reallocation of organizational resources.
Personal Empowerment
One of the key outcomes of cloud computing is the ability to lower barriers to use
and ultimately empower individuals through access to enterprise-scale technol-
ogy. By both lowering the cost and moving to a subscription model from a capital
investment model, cloud providers are enabling everyday users to enter the IT
market. Amazon EC2, for example, offers microserver instances for only two cents
per hour of CPU time as well as high memory and CPU instances. When com-
bined with auto-scaling services, the EC2 model enables IT departments to allo-
cate computing resources as necessary and without a large capital investment.
There are also PaaS and SaaS style services such as Omeka.net (http://omeka.net)
that offer on-demand digital publishing services. While these services are not free,
they enable small libraries to leverage the expertise and IT support of much larger
organizations at relatively low cost. 43
Redefine IT Issues
Second, cloud computing has the ability to redefine key IT issues. For example,
security, stability, and data integrity are important aspects of any IT service
approach. By addressing these issues, and by further allowing organizations to
purchase these services, cloud computing enables organizations to deploy services
with the same or better SLA levels that would have been made available in local
computing environments. Furthermore, cloud computing offers the ability to fun-
damentally change how IT providers approach technology issues such as stability
and data integrity. By using virtualization and redundancy-focused technologies,
cloud service providers can provide advanced versions of these services that would
be cost prohibitive for end users to implement on their own. Likewise traditional
issues with capital allocation and the challenges of responding to rapidly changing
user needs are fundamentally redefined in a cloud environment. By removing the
need for large capital investment and by allowing IT departments to publish IT
resources as services rather than commodities, cloud computing providers facili-
tate the introduction of new IT services outside of formal capital request cycles. As
a result, IT organizations can begin approaching the provision of IT service from
a user-centric perspective.
51046_ch04_p037-046.qxd 5/19/11 12:48 PM Page 44
Conclusion
This chapter has taken a very positive view of the role and potential impact of
cloud computing in libraries and other information organizations. To be sure,
potential pitfalls such as security, privacy, information ownership, resource access,
and IT sustainability are significant issues facing cloud adopters. Rather than con-
sidering any one of these issues to be insurmountable, organizations should con-
sider the decision-making and resource allocation models discussed here to help
guide decisions surrounding cloud computing. As cloud platforms mature,
providers are becoming more accustomed to providing traditional ITSM services,
44 including the definition of SLAs and implementation of ITSM procedures, to
ensure that security and service levels meet the needs of the subscribing organiza-
tion. In each of these cases it is important to balance the risk against the service
need and to select appropriate providers and service agreements based on criteria
relevant to the subscriber.
References
Anderson, Janna, and Lee Rainie. 2010. “The Future of Cloud Computing.” Pew Research
Center’s Internet & American Life Project. June 11. http://www.pewinternet.org/
Reports/2010/The-future-of-cloud-computing.aspx.
Blowers, H. 2010. “From Realities to Values: A Strategy Framework for Digital Natives.”
Computers in Libraries 30, no. 4 (May): 6–10.
Grossman, Lev. 2006. “Time’s Person of the Year: You.” Time, December 13.
http://www.time.com/time/magazine/article/0,9171,1569514,00.html.
Haas, Hugo, and Allen Brown. 2004. “Web Services Glossary.” World Wide Web Con-
sortium. http://www.w3.org/TR/ws-gloss/.
Horrigan, John. 2008. “Use of Cloud Computing Applications and Services.” Pew
Research Center’s Internet & American Life Project. http://www.pewinternet.org/
Reports/2008/Use-of-Cloud-Computing-Applications-and-Services.aspx.
51046_ch04_p037-046.qxd 5/19/11 12:48 PM Page 45
Intel. 2010. “Moore’s Law and Intel Innovation.” Intel. Accessed November 20.
http://www.intel.com/about/companyinfo/museum/exhibits/moore.htm.
Kinney, Bo. 2010. “The Internet, Public Libraries, and the Digital Divide.” Public Library
Quarterly 29, no. 2 (April): 104–161.
Low, Montana. 2010. “Google Is Eating Microsoft’s Lunch, One Tasty Bite at a Time.”
Rescuetime (blog), June 17. http://blog.rescuetime.com/2010/06/17/google-is-eating-
microsofts-lunch-one-tasty-bite-at-a-time/.
Moore, Geoffrey. 2010. “Atmosphere: Core, Content and the Cloud.” YouTube video,
42:33, posted by EVENTS@Google, April 12. http://www.youtube.com/
eventsatgoogle#p/u/8/0swJCYLH2Ck.
Perry, Anali. 2010. The Library Channel. Accessed September 10. http://
lib.asu.edu/librarychannel.
Rowe, Theresa. 2010. “Spotlight on Cloud Computing: Impact! EDUCAUSE Live!”
(presentation). EDUCAUSE. September 17. http://net.educause.edu/LIVE1026.
Smith, Aaron. 2010. “Mobile Access 2010. Pew Internet & American Life Project.”
Pew Internet. July 7. http://www.pewinternet.org/Reports/2010/Mobile-Access-2010/
Summary-of-Findings.aspx.
Smith, Ivan. 2010. Cost of Hard Drive Space. Updated December 3. http://ns1758.ca/
winch/winchest.html.
Smith, Shannon D., and Judith B. Caruso. 2010. ECAR Study of Undergraduate
Students and Information Technology, 2010. ECAR Study of Undergraduate
Students and Technology. October 22. http://www.educause.edu/Resources/
ECARStudyofUndergraduateStuden/217333.
Time. 1983. “MACHINE OF THE YEAR 1982: The Computer Moves In.” October 5.
http://www.time.com/time/magazine/article/0,9171,952176,00.html. 45
ZSR Library (Z. Smith Reynolds Library). 2010. Toolkit. June 2. http://
zsr.wfu.edu/toolkit/.
51046_ch04_p037-046.qxd 5/19/11 12:48 PM Page 46
51046_ch05_p047-058.qxd 5/19/11 12:48 PM Page 47
Chapter 5
Introduction
Overview and Benefits of Cloud Computing
According to The Quotations Page website (http://www.quotationspage.com), it
was Arthur C. Clarke who said, “Any sufficiently advanced technology is indistin-
guishable from magic.” To listen to all the current claims being made about cloud
computing it might easily qualify as magic under Clarke’s definition. The claims
47
have gotten so overblown that the Gartner group noted on October 7, 2010, that
cloud computing, along with many other new technologies, has moved into the
“Peak of Inflated Expectations” (Gartner, 2010). In this chapter, the goal is to pro-
vide you with an understanding of the realities of cloud computing from the point
of view of someone who is both a librarian and a vendor.
So let’s start with a reality. Cloud computing is really not new, and it is not
magic. It is at best an evolutionary step in a long line of computing products and
services that can be utilized by libraries. Let’s put that evolution in perspective.
Mainframe computer technology in the 1960s was a clear early ancestor of cloud
computing, as it was based on a centralized server with remote users. This tech-
nology was later used to support travel reservation booking systems, point-of-sale
systems, and numerous other examples that started in the 1970s and exist yet
today. There have also been e-mail hosted systems for a long time. Remember
AOL™ and Compuserve™? These too were hosted wide-area network-based
applications. Even for libraries, library automation vendors and cooperatives have
long offered hosted systems. OCLC originated in 1967 when 54 Ohio college and
university libraries formed the Ohio College Library Center to develop a cooper-
ative, computerized regional library network. OCLC’s bibliographic database, the
Online Union Catalog, began its operation in 1971. So the basic concepts of cloud
computing are really not new.
51046_ch05_p047-058.qxd 5/19/11 12:48 PM Page 48
At a practical and entirely logical level, cloud computing also offers both the
vendor and library some distinct advantages over the existing environments. These
include:
1. Platform neutrality. The history of hardware vendors is nearly as con-
voluted as the surface of the human brain. They come, they go, they
acquire and divest. As it happens, so do operating systems. For the soft-
ware vendors, delivering application software products into this envi-
ronment is equally complex. It increases the number of quality
assurance tests that must be run, the number of staff needed to program
and support applications, and overall time needed to get a new or
updated application to library customers. The use of a cloud computing–
based approach allows the vendor instead to consolidate, over time, the
number of platforms to be supported and thus will allow them to be as
focused as the customers on being efficient and effective.
2. Reducing overall costs. This is a benefit that most librarians and ven-
dors understand the cloud computing environment is intended to
achieve. For libraries this occurs because the cost of provisioning hard-
ware and the staff to run it can be redirected to the vendor, who through
having a larger base over which to spread those costs can do it in a more
efficient manner and at a lower cost to the library. In addition, because
libraries need purchase only computing services to meet their needs
48 and can quickly scale those services, cost efficiencies are realized there
as well. For the vendors, it means gaining the maximum use out of the
investment they make in hardware through optimal use of talented staff,
thus gaining the greatest efficiency.
3. Ability to refocus library and/or IT staff and redirect cost of ser-
vices to more important value-added services. “Cloud computing
offers an economic advantage by allowing institutions to focus more
resources on differentiating value” (Oblinger, 2010: 4). In discussions
with librarians about cloud computing, they’ll frequently mention that
they do not see that they are adding substantial value to the running of
computing infrastructure. In times of economic crisis, this becomes a
far more important factor, and thus library directors seek to redeploy
staff in order to create substantial value-add for the end user and the
library.
4. Better support. It is natural for libraries to be concerned that moving
support to their vendor may introduce room for new unknown vari-
ables over what they have today. In doing so, however, consider that the
vendors will, just like the libraries, be able to take existing staff and
better focus and train them as a result of the consolidation of platforms
to be supported.
51046_ch05_p047-058.qxd 5/19/11 12:48 PM Page 49
As a result, libraries and vendors both will realize some major new benefits
under cloud computing. In addition the utilization of this technology will enable
libraries to develop new services and capabilities, some of which are discussed at
the end of this chapter in “The Future of Cloud Computing.”
These kinds of features resulted in Gartner (2010) going on to say in its
Hype Cycle Special Report that it believes cloud computing will have a “signif-
icant impact.” That impact will be quite measurable. A recent projection stated:
“by 2012, 80 percent of Fortune 1000 companies will pay for some cloud com-
puting service, and 30 percent of them will pay for cloud computing infrastruc-
ture” (Rhoton, 2009: Chapter 1). Even today, many library automation vendors
are offering and supporting many customers at some level of the cloud comput-
ing environment. Out of over 9,000 product installations that Ex Libris has
today, nearly 1,500 are utilizing some version of the cloud computing environ-
ment. In fact, all of the major integrated library system (ILS) vendors already
offer one model of cloud computing, that is, software as a service (SaaS) solu-
tions.
Moreover, there are efforts underway at this writing that will bring libraries
even further into the cloud computing environment. These include Ex Libris’s
next-generation library services framework, Alma, and OCLC’s Web Scale Man-
agement Services. Each offers new and radically different capabilities based on the
cloud computing model. One attribute in both models is the use of Software
Object Architecture (SOA). “Cloud computing and SOA are different concepts,
50 but they are related. SOA is a pattern of architecture, whereas cloud computing is
an instance of architecture” (Linthicum, 2009: Chapter 2). Ultimately, this archi-
tecture will become an important part of the evolution of cloud computing and
will substantially contribute to the benefits derived from it.
offering. Other examples would be SaaS offerings of open source ILS products
using Amazon EC2 services or some of the aggregated indexes of articles that are
currently available to libraries.
As librarians, it is important to know the configurations being utilized in a
cloud computing–based product offering. As with anything, there are benefits and
potential complications that need to be understood as a result of a multilayered,
multiprovider solution. Asking detailed questions is key. What are the offering
options? What components are based where? What companies will provide valu-
able information concerning what can be expected from the service level agree-
ments (SLAs) as well as the overall scalability and availability of the offering being
offered? For instance, if a supplier uses Amazon EC2 services to provide SaaS
solutions such as an ILS or a data service, it would virtually ensure that scalability,
redundancy, and availability are extremely high. At the same time, it would be
important to understand how the SLA provided and signed by the library’s solu-
tion provider also ensures that Amazon EC2 will meet or support the terms of that
SLA. Likewise, if the solution provider says it provides the cloud-based solution
from top to bottom, the library staff should probe into whether or not they truly
have the same levels of scalability, redundancy, and availability as that available
from firms that focus on those aspects.
Before leaving this overview of cloud computing and why it’s important, it is
equally important to remember that “[c]loud computing is anything but a
mature technology” (Rhoton, 2009: Chapter 29). Consequently, there are
numerous areas where librarians and vendors will need to work together to bring 51
that maturity to the technology. You’ll find a list of these areas in a later section
of this chapter.
Security
Librarians worry about security of cloud computing applications for several rea-
sons. All of these should be examined and addressed as part of the decision to
move to the cloud. Most of these concerns seem to center around the library’s
data, the organizations holding it in the cloud, and its physical storage and safety as
well as whether the security is sufficient to prevent unauthorized access by third
parties.
51046_ch05_p047-058.qxd 5/19/11 3:11 PM Page 52
A frequent question is what happens to the library’s data if its vendor is bought
and/or goes into bankruptcy?
This can be complicated because, in the library field, many times it is thought that nonprofit
entities, if they offer a comparative, cloud-based service, give the customer extra security in
this realm. There may be a belief that there is less risk of it going through organizational
changes that would create substantial likelihood of these types of risks being encountered.
But the comparison is more complicated than that. A comparison of the organizational entities
must include questions about how many products/services have been purchased and/or
launched and then subsequently abandoned or otherwise then ceased to be offered to the
market in a relatively short time frame.
The comparison must also include the financials of all the organizations being consid-
ered, to ensure financial viability in terms of revenue growth and in terms of comparing prof-
itability (those that are for-profit) against revenue in excess of costs for a nonprofit entity. All
organizations bring some level of risk to this consideration. Measuring those risks and making
informed choices based on those measures is the only responsible way to proceed.
Another area of major concern is the privacy of the patron data. The reasons
are obvious, as exposure of patron data causes major uproars and public embar-
rassment for the institution and may well violate laws in cases where FERPA
(Family Educational Rights and Privacy Act) compliance is mandated. So ensur-
52 ing the vendor will handle the patron data appropriately is essential. From the
vendor’s point of view, this can be further complicated if the company operates
on a global basis. Depending on where the data resides in the cloud, laws such as
the USA PATRIOT Act (Uniting and Strengthening America by Providing
Appropriate Tools Required to Intercept and Obstruct Terrorism Act, or
“Patriot Act”) may mean organizations not based in the United States will not
want their data stored on systems that would fall under the reach of the Patriot
Act. For the vendor, this means co-locating data centers in multiple countries,
usually on different continents, so as to also address redundancy and availability
requirements at the same time.
When it comes to measuring the risk to the data itself, there are audits avail-
able that can be used by both librarians and vendors in order to provide assur-
ance that a responsible and well-managed operation is being run. There are a
number of existing security audits, such as SAS70 and ISO/IEC 27001, that
can be used by cloud computing providers, and asking to see those is one step to
take (although providers may want only to indicate if they’ve passed those
audits, as disclosing the details in full can reveal details of their security setup
they may feel would compromise their secure environment). In addition, “[t]he
Trusted Cloud Initiative, the product of an alliance between the CSA (Cloud
51046_ch05_p047-058.qxd 5/19/11 12:48 PM Page 53
Security Alliance) and Novell, will be a third-party, vendor neutral standard for
cloud security and compliance that’s designed to offer additional security certi-
fication requirements for cloud providers as well as educational tools” (Colaner,
2010: 22–23).
Control
Librarians need to be concerned that when moving to the cloud they maintain
enough control to support the “value-add” of their library and services within the
final solution. Some call this the ability to highlight the unique while integrating
the common. Data when moved to the cloud where it becomes a service inherits
the disadvantage of becoming more of a commodity. Commodities become widely
available and typically at a low cost. So, this forces librarians to think where they’ll
add value to that data in order to provide differentiation. This can be done with
cloud computing–based systems via:
• Software configuration options. While it is important that software can
be quickly installed and put into production quickly, librarians seek con-
figurability of the system in order to meet unique needs. This is a critical
way to provide differentiation to their end users. This can be supported
in numerous ways ranging from the data that can be loaded (can the
library load local and unique data?) through to the software options (for
instance, can the library configure relevance ranking on all data, facets
used, displays, etc?). For vendors, trying to provide the same level of 53
flexibility systems had when they existed in individual installations can
prove difficult. In the end, the new benefits derived from this architec-
ture need to be balanced against the loss of configuration options to
make an informed decision on which course is best for the library.
• Software extensions. Vendors can find it hard to provide enough configu-
ration options to make the system, using shared code and data, flexible
enough to meet all the needs of the customer base. In order to do this,
extensions and the ability to configure those extensions will become impor-
tant. Open platforms are one model, and others address the same concern.
Yet another area of control that is of concern to both librarians and vendors is
recognizing that many customers innovate on a regular basis and may want access
to their data at levels that would permit that innovation. Providing this can prove
challenging for the vendors while simultaneously maintaining a secure and stable
environment for their other customers.
In analyzing the concerns surrounding control, even though libraries are
clearly more comfortable with products and data in the cloud than they used to be,
the final decision on using cloud solutions will hinge on deciding if the business
needs clash with the unique materials of the institution. In the end, it will require a
51046_ch05_p047-058.qxd 5/19/11 12:48 PM Page 54
very institutional decision, and the library administration must be sure they are
philosophically comfortable with that decision.
Branding
As more and more data moves into the cloud and the software using that data
becomes cloud based, it becomes harder and harder for the end user to know
exactly where the information appearing on the screens is originating. As this hap-
pens, libraries will want and need to brand the information they serve. By so
doing, it will establish both the authority and authenticity of the information and at
the same time remind the end user of the value of the library in delivering appro-
priate information to meet their needs. Branding may be done via logo displays
or, depending on the type of data object being delivered, style sheets and/or plac-
ing the logo in the image or background behind the information display. There
are no easy answers here, but clearly this issue is becoming one of increasing
importance and one where smart and creative solutions will quickly gain wide
market acceptance.
Service
Both librarians and vendors realize that for librarians to turn additional
service/support over to the vendor, an element of trust is needed. This should be
backed up by an SLA that details what the service expectations will be, how they
54 will be met, and in what time frames. This is fair and good, as expectations should
be clearly documented by both parties to ensure that the measurements are fairly
applied and evaluated.
One area of service that is hard for both the vendor and librarians to control is
that of network/bandwidth availability. Most cloud-based solutions will not be
based on private networks, and, as such, a backhoe or competing services on the
network can have a major impact on overall system performance. As one librarian
noted about bandwidth concerns: “It’s hard to compete with YouTube on
campus,” and, of course, this is a very valid point. There are technologies available,
however, that can be used to protect bandwidth for important and major functions
on the campus, and their use may become necessary in the future.
Additional Concerns
In the final analysis of advantages and disadvantages, “the most important consid-
eration is not whether a potential solution satisfies the definition of cloud com-
puting but rather whether it adds value” (Rhoton, 2009: Chapter 1). Because of
the requirement of libraries to be far more efficient and effective with their
resources in the current economic environment, the cloud computing model, com-
bined with new functional capabilities that derive from both the architecture and
the software capabilities developed, has the potential to add substantial new value.
51046_ch05_p047-058.qxd 5/19/11 12:48 PM Page 55
Legislation
A more recent trend that is troubling to both librarians and vendors is the increas-
ing intervention by local, state, and/or federal governments to put into place legis-
lation with the intent to protect user data. An unfortunate, and likely not well
understood, consequence of this legislation could be a greatly slowed down adop-
tion rate of cloud computing while these issues are sorted out in the marketplace
and courts.
In the United States, the Patriot Act is widely seen as an issue in the adoption
of cloud computing solutions. The government’s ability to access data without the
data owner even being informed will cause many organizations, particularly cor-
porate libraries, to take a more cautious approach wherein data is stored on
56 machines owned and controlled by the organization.
The New York Times reported on September 19, 2010, that “cloud-based
breakthroughs face a formidable obstacle in Europe, however: strict privacy laws
place rigid limits on the movement of information beyond the borders of the 27-
country European Union” (New York Times, 2010).
There will likely be other examples of this type in the future, and this kind of
prescriptive legislation is the result of legislators not fully understanding that
which they’re trying to legislate. As a result, in the end, they’ll cripple their con-
stituents. Other countries, unhindered by these inflexible approaches, will enjoy
the benefits of cloud computing earlier and will be more agile and will move ahead
in what is a highly competitive global environment. Neither libraries nor the edu-
cators they serve can afford this risk. It would be far wiser for libraries and vendors
to proactively and jointly educate administrators and legislators about what can
be done today that addresses security issues so that better and more appropriate
legislation results.
and the future will be very dependent on the many internal and external dynamic
factors discussed in this chapter, including social networking, analytic capabili-
ties, and standards and legislative environments.
Even so, given cloud computing today, there are several things likely in the not-
too-distant future. Through the use of data clouds, cloud computing, analytics,
and mathematics, the value of librarianship will be turned into a new vehicle of
information processing that can be readily and widely deployed across the web in
a scalable fashion. “There is a growing need for filtering, sorting, categorization
and analytics to help users manage the vast repositories of both structured and
unstructured information and to easily find and extract the information that is of
interest” (Rhoton, 2009: Chapter 29). As a result, cloud computing has the poten-
tial to become a vehicle for the reinvigoration of librarianship.
Clouds will achieve interoperability, not only between like systems, but
between all types of systems, and this will lead to a more seamless integration of
library services into numerous environments such as course management, online
education, and certainly worker re-education and training programs.
On the flip side of these positives, librarians need, as a group, to realize they
should “expect some demographic effects of cloud computing. By virtue of its loca-
tion independence there may be increases in off-shoring. There may also be impact on
employment as workers need to re-skill to focus on new technologies and business
models” (Rhoton, 2009: Chapter 1). As services move into the cloud and become
deployable across the web, this is a very real potential risk. As with so much of tech-
nology today, the new, well-paying, and lasting jobs to be created will require that 57
workers in the librarian profession engage in getting up-to-date training and education
and to employ that education in developing leading edge information services.
In a recent issue of EDUCAUSE Review, David Lewis, Dean of the IUPUI
University Library, described the future of libraries this way:
Ten years from now, the historic corpus of printed books will likely have been con-
verted to digital files. . . . Print copies will be stored in long-term print repositories.
. . . Ten years from now, digital book readers will be common, and print-on-
demand machines will be better and cheaper. . . . Finally, ten years from now, every-
one will expect that all documents should be instantly available anywhere and in all
the forms—digital or paper—that might be useful. (Lewis, 2010: 11)
This is indeed a possible scenario. It is one viable way for libraries to benefit and
to be part of the fabric of that coming environment if they’re operating in a cloud
computing environment.
This chapter started with a quote by Arthur C. Clarke and so it will conclude
with another. According to The Quotations Page (http://www.quotationspage.com),
in an address to the U.S. Congress in 1975, Clarke said “I’m sure we would not
have had men on the Moon if it had not been for Wells and Verne and the people
who write about this and made people think about it.”
51046_ch05_p047-058.qxd 5/19/11 12:48 PM Page 58
Hopefully this chapter will contribute to librarians thinking about cloud com-
puting and leading libraries toward a future that utilizes and benefits from the
same.
References
Colaner, Seth. 2010. “Cloud Computing Security Considerations.” PC Today 8, no. 9:
22–23.
Gartner. 2010. “Gartner’s 2010 Hype Cycle Special Report Evaluates Maturity of 1,800
Technologies.” Gartner, Inc. October 7. http://www.gartner.com/it/page.jsp
?id=1447613.
Lewis, David. 2010. “The User-Driven Purchase Giveaway Library.” EDUCAUSE Review
45, no. 5: 10–11.
Linthicum, David. 2009. Cloud Computing and SOA Convergence in Your Enterprise: A
Step-by-Step Guide. Reading, MA: Addison-Wesley Professional. Kindle e-book.
New York Times. 2010. “Cloud Computing Hits Snag in Europe.” New York Times, Sep-
tember 19. http://www.nytimes.com/2010/09/20/technology/20cloud.html.
Oblinger, Diana. 2010. “Stewards for Higher Education: Looking at Clouds and the Top-
Ten IT Issues.” EDUCAUSE Review 45, no. 3 (May/June): 4.
Rhoton, John. 2009. Cloud Computing Explained: Implementation Handbook for Enter-
prises. Recursive Press. Kindle e-book.
58
51046_ch06_p059-068.qxd 5/19/11 12:49 PM Page 59
Chapter 6
Cloud Computing
for LIS Education
Christinger R. Tomer and Susan W. Alman
Introduction
The future of libraries and archives may well be determined to a significant degree
by the technological competence of the librarians and archivists who run them.
Cloud computing and its capabilities provide educational programs in library and
information science new and potentially powerful choices in supporting the tech-
nological components of their programs. There is a need for Library and Infor-
mation Science (LIS) programs to develop a consortium using cloud computing as
an integrating and sharing mechanism to build a virtual learning and computing
laboratory. 59
The technological component of LIS education has long been constrained by
the high costs of information technologies and institutional priorities that tend to
discourage experimentation and emphasize administrative concerns such as secu-
rity. Notwithstanding a general increase in the attention devoted to information
technology in basic courses, the vast majority of LIS students enrolled in ALA-
accredited programs learn relatively little about library and archival systems, such
as integrated online library systems or digital repositories, and even fewer of them
acquire hands-on experience with these or other relevant systems.
This technological deficit is a significant, if not readily acknowledged prob-
lem of LIS education and one that may ultimately threaten many programs, given
the growing importance of technology-oriented skills in the professional market-
place. Cloud computing alone may not constitute a solution to the problem—the
acuity and technical competencies of LIS faculty are presumably the most impor-
tant factors in determining the technological competence of future librarians and
archivists. The potentially favorable economies and adaptability of the cloud com-
puting environment, particularly when combined with what has been learned in
other disciplines through experiments in the development of virtual laboratory
capabilities, offer opportunities for LIS programs to render major improvements in
51046_ch06_p059-068.qxd 5/19/11 12:49 PM Page 60
the technical infrastructure supporting teaching and learning about learning tech-
nologies, in consortial or independent modalities.
Owing to the capabilities of cloud computing, LIS programs have new and
potentially powerful choices in supporting the technology components of their
educational programs. The aim of this chapter is to identify and explore briefly
the opportunities afforded by cloud computing, culminating in a series of specific
recommendations for the support of LIS education.
sharing. Perhaps most important of all, the flexibility of the cloud computing model
and the ease with which virtual machines can be built allows course designers to
build systems that are designed to meet the requirements of specific courses (Seay
and Tucker, 2010). For example, at North Carolina State University, the combined
use of cloud computing and virtualization has enabled the university to expand the
number of the production images available for use from approximately 20 to more
than 600, thereby vastly increasing the array of resources and configurations avail-
able to faculty and students (Shaffer et al., 2009).
The idea of virtual laboratories is not unknown in LIS education. While it does
not fully qualify in technical terms, distributed access to the Internet Public Library
(IPL) by various LIS education programs for instruction and training has set an
important precedent, whereby collaborative technologies are used to increase the
array of resources available for teaching and learning and provide shared access to
them. The IPL (http://www.ipl.org) originated in a graduate seminar at the Uni-
versity of Michigan in 1995. What became the IPL was defined at the outset by
two ideas: to ask some interesting and important questions about the interconnec-
tions of libraries, librarians, and librarianship with a distributed networked envi-
ronment, and to learn a lot about these issues by actually designing and building an
Internet-based reference resource/service. In 2006, responsibility for the IPL was
transferred to a consortium that is led by Drexel University, with the University of
Michigan and Florida State University as major contributing partners.
The precedent established by the IPL’s “virtual learning laboratory” has been
reinforced lately by LibLime and OCLC, each of which has used cloud computing 63
and virtual machine technologies to provide LIS faculty and students with access
to archetypical systems. In the case of LibLime, it has created and maintains
instances of its enhanced version of Koha, the open source integrated library
system (ILS), for use in LIS education programs, whereas OCLC provides access
to CONTENTdm, its digital archiving platform, as part of a broader initiative in
support of LIS education. See LibLime (http://www.liblime.com/), Koha
(http://koha.org/), and Library and Information Science Education Program
(http://www.oclc.org/ca/en/community/education/lis/) for more information.
The significance of access to what are effectively educational versions of CON-
TENTdm and Koha is that for the first time LIS students in significant numbers have
the opportunity to work with and learn about systems that are effectively representa-
tive of the platforms that define digital libraries and archives. In addition, virtual
machine images of other key LIS systems, such as DSpace, Evergreen, and Islandora,
are now available for downloading and installation. This means that key components
of a virtual learning laboratory providing access to a series of relevant systems and
resources for LIS education are already available, and it is now mainly a matter of
how and under what specific circumstances such a laboratory should be built.
There are plenty of options available. LIS programs could band together using
cloud-based services as an integrating and sharing mechanism and build a virtual
51046_ch06_p059-068.qxd 5/19/11 12:49 PM Page 64
computing and learning laboratory designed to serve the requirements of the par-
ticipating programs, much like the IPL is shared by the programs that underwrite
its ongoing costs. The programs participating in such a consortium would still
have the option to supplement these shared services through private clouds and/or
on-site systems, or they could pursue the development of a virtual laboratory capa-
bility on an independent basis.
In any of these instances, it seems reasonable to imagine that students in par-
ticular would have access to computing and networking resources not available to
them on the basis of locally provided and traditionally configured computing ser-
vices, that the technological aspects of LIS education would be substantially, if not
uniformly improved, and that continuities without parallel in LIS education could
be achieved. For example, students in a course on digital libraries could work with
digital archiving systems like Archon and ICA-AtoM to learn about design, inter-
operability, and administrative requirements of those systems, whereas students
in a course on metadata might work with the same systems in order to achieve an
operational understanding of how archivists describe digital objects. Of equal
importance, a more uniform approach to technology education might establish a
meaningful basis for defining and assessing technological competency within the
library and archival professions.
The key issue in building a virtual lab is selecting the systems and applications
that will be supported. Because many vendors have been reluctant to allow their
systems and applications to be released for educational use—OCLC is an obvi-
64 ous exception—most of the options entail the use of open source software. But
there are decided benefits to working with library and archival systems that have
been developed in the open source environment. Licensing issues and compli-
ance requirements tend to be minimal, which means, for example, that systems
such as EPrints or DSpace or Koha can be built and configured as virtual
machines and then deployed under a variety of circumstances without the con-
sent of the developers.
Another significant benefit is that the open source systems that have been
developed by and for the library and archival communities are typically based on
either the so-called “LAMP” architecture—Linux, Apache, MySQL, and PHP—or
a Java-based framework often incorporating Apache Tomcat. Those architectures
are sufficiently standardized to make installation, configuration, and deployment
comparatively simple tasks (Morelli and de Lanerolle, 2009). For example, in the
domain of digital archiving, Archon and ICA-AtoM are based on the LAMP archi-
tecture. EPrints, probably the most widely used of the digital repository systems, is
also LAMP based, whereas DSpace and Fedora Commons, the other prominent
digital repository systems, are based on Java and Apache Tomcat. See Archon
(http://www.archon.org/), ICA-AtoM (http://ica-atom.org/), EPrints (http://
www.eprints.org/), DSpace (http://www.dspace.org/), and Fedora Commons
(http://fedora-commons.org/) for more information.
51046_ch06_p059-068.qxd 5/19/11 12:49 PM Page 65
Plone Content management system built on top of the open source application server
Zope and an accompanying content management framework. http://plone.org/
51046_ch06_p059-068.qxd 5/19/11 12:49 PM Page 66
Archivists’ Archival data management system to provide broad, integrated support for
Toolkit the management of archives. http://www.archiviststoolkit.org/
Archon So-called simple archival information system supporting EAD and MARC.
http://www.archon.org/
Greenstone Suite of software tools for building and distributing digital library
collections. http://www.greenstone.org/
Islandora Combines the Drupal and Fedora software applications to create a digital
asset management system with support for collaboration at several levels,
including metadata and narration. http://islandora.ca
Koha Koha was the first open source integrated library system. It was created in
1999 by Katipo Communications for the Horowhenua Library Trust in New
Zealand. http://koha.org/
Omeka Web publishing platform for the display of library, museum, archives, and
scholarly collections and exhibitions, combining features of content
management, collections management, and archival digital collections
systems. http://omeka.org/
SYSTEM DESCRIPTION/URL
Open Conference An open source conference management system that can create
Systems and manage the complete web presence for an academic
conference. http://pkp.sfu.ca/?q=ocs
Open Journal Systems An open source journal management and publishing system that’s
purpose is to make open access publishing a viable option for
more journals. http://pkp.sfu.ca/?q=ojs
Open Harvester An open source metadata indexing system that allows you to
Systems create a searchable index of the metadata from Open Archives
Initiative (OAI)–compliant archives. http://pkp.sfu.ca/?q=harvester
Conclusion
The future of librarianship and the archival profession may well be defined to a
significant degree by the technological competence of librarians and archivists.
But that future is at issue, in part because LIS educators have commonly assumed
that it was sufficient for the technological component of LIS education to engen-
der awareness, as opposed to detailed, substantive knowledge and proficiency. In
reality, librarians need to master the technologies that define the twenty-first-cen-
tury library in order to exert appropriate controls and use those technologies in
ways that clearly and creatively serve the interests and needs of library users. 67
Moving into the cloud and building the virtual learning laboratories that provide
future librarians with access to the technologies they will use and need to under-
stand is an important step in that direction and one that LIS education can ill-
afford to forego.
References
Aburdene, M.F., E.J. Mastascusa, and R. Massengale. 1991. “A Proposal for a Remotely
Shared Control Systems Laboratory.” In Frontiers in Education Conference, 589–592.
Twenty-First Annual Conference—Engineering Education in a New World Order Pro-
ceeding, West Lafayette, Indiana.
Aharony, Noa. 2008. “Web 2.0 in U.S. LIS Schools: Are They Missing the Boat?”
Ariadne 54. http://www.ariadne.ac.uk/issue54/aharony/.
Koretsky, Milo D., Danielle Amatore, Connelly Barnes, and Sho Kimura. 2008. “Enhance-
ment of Student Learning in Experimental Design Using a Virtual Laboratory.” IEEE
Transactions on Education 51 (February): 76–77.
Malinconico, S.M. 1992. “What Librarians Need to Know to Survive in an Age of Tech-
nology.” Journal of Education for Library and Information Science 33: 226, 228–232.
Morelli, Ralph, and Trishan de Lanerolle. 2009. “FOSS 101: Engaging Introductory Stu-
dents in the Open Source Movement.” In Proceedings of the 40th ACM SIGSCE Tech-
nical Symposium on Computer Science Education (Chattanooga, Tennessee, March
4–7), 311–315. New York: ACM Press.
51046_ch06_p059-068.qxd 5/19/11 12:49 PM Page 68
Scott, Dan. 2010. “Easing Gently into OpenSRF, Part 1; and Easing Gently into Open-
SRF, Part 2.” Code{4}Lib Journal 10 (June). http://journal.code4lib.org/articles/3284
and http://journal.code4lib.org/articles/3365.
Seay, Cameron, and Gary Tucker. 2010. “Virtual Computing Initiative at a Small Public
University.” Communications of the ACM 53 (March): 75–83.
Shaffer, Henry E., Samuel F. Averitt, Marc I. Hoit, Aaron Peeler, Eric D. Sills, and Mladen
A. Vouk. 2009. “NCSU’s Virtual Computing Lab: A Cloud Computing Solution.”
Computer 42, no. 7: 94–97. doi:10.1109/MC.2009.230.
68
51046_ch07_p069-086.qxd 5/19/11 12:49 PM Page 69
Part II
TECHNOLOGIES
51046_ch07_p069-086.qxd 5/19/11 12:49 PM Page 70
51046_ch07_p069-086.qxd 5/19/11 12:49 PM Page 71
Chapter 7
Introduction
In recent years libraries have turned to a new generation of interfaces more in tune
with the expectations of today’s users shaped by their experiences of the current
web. Against the context of incredibly powerful search tools, intuitive navigation,
and socially engaging and visually appealing virtual destinations on the web, the
incumbent generation of online catalogs falls short in the way that they present
the collections and services of the library. The past five years have seen a continual
advancement of new products, both proprietary and open source, working toward
ever more modern user-friendly features and broader scope of search. These new 71
discovery services have become increasingly expansive and virtual, tapping ever
more deeply into the concepts and technologies of cloud computing.
In this chapter, the term “discovery services” is used to describe this new genre
of end-user library interfaces. Other terms commonly associated with this genre
include “next-generation library catalogs” and “discovery interfaces.” The term
“discovery services” is applied generally, including services installed on the
library’s own servers and those offered through software as a service arrangements.
The concept of discovery services isn’t brand new. These products have been
available in some form since about 2004. Libraries can consider these discovery
interfaces not as bleeding-edge technologies but as relatively well-established and
maturing tools that have steadily improved the interfaces offered to their users.
The initial, ground-level round of products, often characterized as next-gener-
ation library catalogs, modernized the interface but remained mostly focused on
local library collections. Subsequent offerings and new product versions have con-
tinually elevated their scope of search, adoption of social networking concepts, and
improved integration with the broader enterprise of library products and services.
The products and services available today fall in a continuum ranging from
those focused more on local content physically held in the library to those that
progressively address the diffuse cloud of content representing a broad view of
51046_ch07_p069-086.qxd 5/19/11 12:49 PM Page 72
library collections that includes the body of all the electronic content libraries con-
sider within their scope of interest. Libraries will differentiate their selection of a
discovery service, or how it’s implemented, based on their strategic approach to
how they want to position their collections to their users.
This new genre of library discovery services aims to go beyond the capabilities
of the traditional online catalog delivered as a module of the integrated library
system (ILS) and to present a more modern interface to library users. These prod-
ucts have evolved to expand the search of scope far beyond the aspects of library
collections managed by the ILS, encompassing materials managed by other sys-
tems. The latest and most ambitious set of products extend the reach of discovery
even further, layering in access to the vast collection of individual articles repre-
sented within a library’s subscriptions to electronic resources. These new prod-
ucts aim to deliver an experience of libraries on the web that reflects a more
user-centric approach and that better represents library collections that include
print, electronic, and rich media.
This chapter aims to provide a general understanding of discovery services,
how they fit into the larger context of library technologies, and some of the trends
that have shaped the current landscape. No attempt is made to review or provide
detailed information on the specifics of the individual products within this genre.
products have been available for a number of years, the majority of libraries con-
tinue to rely on traditional online catalogs. We can expect the proportion of
libraries implementing discovery services to increase over time. Contributing fac-
tors include the lack of ongoing development of ILS online catalog modules,
aggressive marketing of discovery services products, bundling of discovery ser-
vices with the ILS, as well as increased interest by libraries in improving their
offerings facing end users. Limiting factors include the lack of financial or person-
nel resources to implement any new technology products, waiting for product
offerings to further mature, and long selection and procurement cycles.
remote servers, and especially the quality and quantity of results returned by each
target resource. These federated search products initially made use of standard
search-and-retrieval protocols such as Z39.50 but eventually expanded to also use
more efficient XML gateways or other technologies that improved the power and
performance possible through this search architecture.
Federated search products popular in the library arena include the following:
MetaLib from Ex Libris (announced July 2000)
Research Pro from Innovative Interfaces
WebFeat (released about 1998)
360 Search from Serials Solutions (released in January 2005 under prod-
uct name Central Search)
MasterKey and Pazpar2 from Index Data, a firm specializing in open
source library software (MasterKey is a complete federated search prod-
uct. Pazpar2 is middleware that can be integrated into other products.)
MuseSearch from MuseGlobal (released 1992; several of the federated
search products rely on technology license from MuseGlobal)
Although federated search products such as these provide some ability to expand
the scope of search for library resources, they did not attempt to supplant the
online catalog of the ILS. Libraries would continue to offer their online catalog
for searching their local collection; federated search utilities generally provide a
way to consolidate search for electronic resources but do not subsume functional-
ity of the online catalog. 75
Although federated search products serve as useful tools in assisting users with
access to electronic content, they are not commonly positioned as replacements
for the online catalog of the ILS. These products paved the way for discovery
interfaces, a more complex genre of products that modernize the traditional online
catalog and expand the scope of search.
quent transfers for newly added content. End users then perform searches against
the discovery service, accessing the target servers only once an article has been
selected. The wholesale and incremental harvesting that supports consolidated
search discovery services can be performed at the convenience of the information
provider during off-peak times. In addition to the benefits associated with more
powerful search capabilities for end users, the consolidated index model offers a
more sustainable technical model for information providers compared to feder-
ated search.
depending on the interests of the user. Most discovery services offer rel-
evancy ranking by default, with options to sort by other categories.
• Faceted navigation. This involves terms and categories presented as
part of a search interface that allows the user to make a series of selec-
tions to incrementally narrow the results to hone in on items of interest.
The facets are generated dynamically from the query results, usually
with the number of items displayed associated with the term or category.
• More intuitive navigation. Discovery interfaces are generally quite
attuned to usability issues, aiming to make dramatic improvements
beyond online catalogs that gained a reputation for unintuitive and
quirky interfaces, designed more for librarians and expert searchers than
for the general public.
• Enhanced visual display. Rather than rely on purely text-oriented
descriptions, discovery services make more frequent use of graphics and
icons, such as cover images of books and DVDs, to improve the visual
appearance of the interface.
• Expanded scope of search. Online catalogs provide access only to the
resources managed by the ILS; discovery services have the capability to
include many other sources of library-related content, such as that from
institutional repositories and collections of digitized images or manu-
scripts. More recent products expand the scope of search to the articles
represented in library subscriptions through large consolidated indexes
of preharvested content. 77
• User-supplied rating and reviews. In tune with Web 2.0 concepts, dis-
covery services allow users to supply reviews, rate materials, and make
recommendations to other users.
Such features are commonplace among the popular nonlibrary web destinations.
The first wave of discovery services helped libraries catch up to expectations of
their users established by their experiences with other information-oriented web-
sites that were sorely lacking in the majority of online catalogs.
service. The process of exporting and re-indexing may involve additional pro-
cessing steps to facilitate the creation of facets, to control relevancy weightings, to
determine presentation, and to make other transformations. In addition to the
basic bibliographic record, other data elements such as location codes, authority
references, and circulation status may also be transferred. Following the initial
wholesale transfer of data from the ILS to the discovery service, frequent updates
will be scheduled to keep the indexes current. How often to perform incremental
updates depends on how rapidly new materials are added to the library’s collec-
tion but in most cases would take place daily.
In addition to the data harvested from the ILS, the discovery interface requires
other information in real time, such as whether an item is on the shelf, checked
out, or on hold for another user. Such status information can be obtained as
needed through a real-time request from the discovery service to the ILS, using
any of a number of techniques, such as through a standard protocol such as NCIP
or SIP2 or through other APIs.
Some of the discovery systems simply rely on the web-based online catalog
when detailed information is needed regarding the current status of an item. The
discovery interface would deep link into the online catalog for a specific item, taking
advantage of its native ability to display its complete status information, as well as
invoke user services such as placing holds. While simpler to accomplish, the tran-
sitions between pages delivered by the discovery service and those of the online
catalog can introduce confusion to the user. A more elegant, though technically
78 complex, approach brings all this functionality directly into the discovery service.
Discovery services include a set of services for users tied to their personal pro-
file. Such features might include the abilities to save search queries, mark items
for later consultation, send alerts of new materials, and set up a profile of prefer-
ences or interests. It may also be possible to link users’ profiles in the discovery
interface to their account in the ILS, making it possible to view items charged,
place holds, renew items, or pay fines within the interface of the discovery service.
This more sophisticated level of integration may not yet be available in all the dis-
covery services, and it’s still common for users to have to deal with the activities
managed by the ILS and those managed by the discovery service separately.
As discovery services began to proliferate, the need emerged to find standard
ways to connect them with ILSs. The Digital Library Federation, now part of the
Council on Library and Information Resources, charged a group to develop a stan-
dard protocol in this arena. The Integrated Library System Discovery Services
Task Group, chaired by John Mark Ockerbloom, was formed in 2007 and devel-
oped a set of protocols that support graduated levels of interoperability. The
lowest of the four levels, dubbed Basic Discovery Interface, uses Open Archives
Initiative Protocol for Metadata Harvesting (OAI-PMH) for transfer of data from
the ILS to the discovery service and passes control back to the online catalog of the
ILS for end-user services. Level 4 describes a robust level of interoperability where
51046_ch07_p069-086.qxd 5/19/11 12:49 PM Page 79
the discovery interface handles all aspects of search and relevant services (Breed-
ing, 2008, a summary of the ILS-DI proceedings and outcomes; Ockerbloom,
2008, the full report of the task group).
Discovery: Mix and Match with the ILS
While the ILS divides its functionality into different modules, they cannot gener-
ally be broken apart. It’s not practical, for example, for a library to describe mate-
rials with the cataloging module from one product, perform circulation functions
with another, and manage the business aspects of procurement through the acqui-
sitions module of yet another. Such a best-of-breed approach of assembling an
automation environment from the best modules has not been possible with the
ILS products offered to date. New products created outside of the bounds of the
ILS, such as OpenURL link resolvers, electronic resource management systems,
and—most recently—discovery services, have followed a strategy of interoperabil-
ity rather than being tied to any specific ILS.
All of the discovery interfaces currently available have been designed to operate
with any of the major ILSs. Libraries may elect to adopt a discovery product from
the company from which they purchased their ILS or from a competitor or imple-
ment one of the open source offerings. The end-user discovery arena allows a mix-
and-match approach that has not previously been a practical alternative with the
other core modules of the ILS.
As an ILS vendor approaches discovery products, they are motivated to create
offerings that will appeal to the libraries that use their own products but also to 79
libraries that use a competing ILS. As interest wanes in traditional online catalogs,
ILS vendors focus their development efforts on the development of discovery ser-
vices instead. These discovery service products are offered to their existing library
customers as add-ons to complement or replace their online catalogs. Discovery
services offered by ILS vendors include the following:
Encore from Innovative Interfaces (introduced May 2006)
Primo from Ex Libris (February 2006)
LS2 PAC from The Library Corporation (October 2008)
Enterprise from SirsiDynix (June 2008)
Visualizer from VTLS (January 2008)
BiblioCommons, launched by Canadian BiblioCommons (2008)
OCLC, a global membership organization, but also involved with a number of ILS
products, began offering WorldCat Local as a discovery service designed to dis-
place the need to operate the online catalog of the library’s ILS.
Another set of discovery services have been created by commercial companies
outside the fold of ILS vendors. These products include the following:
AquaBrowser Library, initially developed by Medialab Solutions of Am-
sterdam, the Netherlands, found use by libraries in Europe beginning
51046_ch07_p069-086.qxd 5/19/11 12:49 PM Page 80
about 2001 and was distributed in North America by The Library Cor-
poration beginning in September 2004. R.R. Bowker acquired Media-
lab Solutions in June 2007; in 2009 control of AquaBrowser shifted to
Serials Solutions, a sister company under common ownership of Cam-
bridge Information Group.
Summon, a discovery service offered by Serials Solutions, was launched in
early 2009. Serials Solutions has been involved in providing a slate of
products surrounding management and access to electronic content but
does not offer an ILS. Serials Solutions offerings now include two discov-
ery services, AquaBrowser Library, with more of a focus on local library
collections, and Summon, which it characterizes as a web-scale discovery
service that also encompasses articles and other electronic content.
In addition to the offerings of these organizations, all offered as proprietary
solutions, open source alternatives have also been created, including these:
VuFind, originally created by the Falvey Library at Villanova University in
2007, has found extensive use by other academic and public libraries
worldwide. This discovery service relies on open source components such
as Apache SOLR and Lucene and uses the PHP programming language.
Blacklight, created at the University of Virginia, finds use by other
libraries, such as Stanford University Libraries, in addition to its home
institution. Blacklight also uses Apache SOLR and Lucene and the
80 Ruby and Rails development framework.
eXtensible Catalog, a project of the River Campus Libraries of the Univer-
sity of Rochester with funding from the Andrew W. Mellon Foundation,
has created a number of tools that facilitate the deployment of discovery
interfaces and is working toward a Drupal-based discovery service.
SOPAC, or the Social OPAC, originally created by John Blyberg at the Darien
Library in Connecticut, emphasizes user interactions, such as reviews, rat-
ings, and other Web 2.0 concepts. The open source Drupal content man-
agement system provides the underlying foundation for SOPAC.
content from that publisher. Such increases in use amplify the value of the con-
tent, providing motivation for content providers to cooperate with discovery ser-
vice providers. Some content providers continue to be reluctant to partner with
discovery services, concerned that wholesale harvesting of their content might
increase the risks of unauthorized access. Content providers also express concerns
that their content loses some degree of branding when accessed through a more
generic discovery service rather than through their own custom interface. Any
problems with the discovery interface may also reflect negatively on the providers
of the content. The growing portfolio of partnerships between discovery service
developers and content providers gives evidence of increased cooperation in this
arena.
In the competition among discovery service products, one of the most impor-
tant differentiating factors involves the completeness of the comprehensive index.
Ideally, each of the discovery services would obtain content from all of the compa-
nies and organizations that provide some sort of information product to libraries.
In reality, however, not all the publishers and providers cooperate with any or all of
the discovery service providers. Some discovery providers have been able to make
arrangements with providers of aggregated databases of articles, gaining access to
very large numbers of articles through a single business partnership. In some cases
the producer of an aggregated information product may choose not to cooperate
with a discovery service. In those cases, the discovery service provider can
approach each of the individual publishers involved to gain access to the articles in
the e-journals covered in an aggregated product. In evaluating the comprehen- 83
siveness of a discovery product, it can be difficult to assess what content is repre-
sented through deals with large aggregators and what has been gained by indexing
content provided by individual publishers.
Aggregated indexes can involve both citation data and full texts of articles.
Many, if not most, of the native interfaces of the information products to which
libraries subscribe support searching on both citations and full text. Google
Scholar indexes the full texts of articles. Given that most search environments
involving journal articles support full-text searching, end users naturally expect
this level of access rather than searching only citation data. Ideally, a discovery ser-
vice based on a central aggregated index would have both high-quality citation
metadata and the full texts of all of the articles from all the publishers and
providers. The current state of the products, however, involves high proportions of
citation-only searching. The technology platforms of the products generally sup-
port full-text searching, but many of the business arrangements between content
and discovery providers currently involve citation data. We can expect higher por-
tions of full-text searching in the discovery services over time.
The evaluation of a discovery service needs to involve an analysis of the con-
tent indexed relative to the library’s current subscriptions. The overall volume of
indexed content may differ from the relative proportion of titles covered to which
51046_ch07_p069-086.qxd 5/19/11 12:49 PM Page 84
the library actually subscribes. Because end-user searching usually will be scoped
to those resources to which the library subscribes, it’s not so much the total
number of articles indexed that matters most, but rather the best coverage of active
subscriptions.
These arrangements with content providers almost always specify that the
content will be exposed only to mutual subscribers. Users taking advantage of a
discovery service will see results from a given resource only if their library sub-
scribes to that resource. An important part of the configuration process of a dis-
covery service for a library involves setting up the profile of subscriptions. This
profile will filter search results to not display results from resources that a user
cannot access.
Unlike online catalogs, which almost always can be searched by the general
public regardless of affiliation with the institution, discovery services may require
authentication in order to display results from a library’s licensed resources. Some
implementations of discovery services prompt all remote users to sign on prior to
searching. Others may allow searching by users who have not signed in but will
show only resources that allow free access, such as the library’s local holdings from
its ILS, open access journals, local repositories, and the like. Through IP-based
authentication, in-library and on-campus computers can be configured to search
licensed resources without users signing on through their personal account.
The creation of the aggregated centralized index also involves a great deal of
technical work on the part of the discovery service provider. This discovery model
84 requires a highly scalable platform capable of loading and indexing many hundreds
of millions of items. The platform will need to index citation and full-text content as
delivered by the information resource providers. In addition to the initial data loads,
the index must be kept up to date with incremental additions from each product
represented. For products including time-critical content, such as current news
sources, new content may need to be added daily, or even more frequently.
The maintenance of a consolidated index for a discovery service represents a
great deal of labor intensive and technically complex work in addition to the busi-
ness negotiations involved in acquiring the content. Given the vast amount of busi-
ness, procedural, and technical work involved, much less the creation of an
appropriate highly scalable technology platform, it is not surprising that only a
few of these discovery services based on consolidated indexes have been created
and that they come out of commercial organizations able to devote significant
resources. While it would be possible for a library or consortium of libraries to
create such an index, so far all of these projects have emerged almost exclusively
from the commercial sector.
These discovery products based on large aggregated indexes also incorporate
the content from the local library. In addition to the content obtained from the
external providers of electronic resources, these indexes also harvest the content
from the ILSs of the libraries that use the service and any other local collections of
51046_ch07_p069-086.qxd 5/19/11 3:14 PM Page 85
interest. This combination of content from local and remote resources results in
the ability to present a single search box that returns unified results that span all
the different aspects of a library’s collection.
This new generation of discovery services based on comprehensive centralized
indexes appeals primarily to research and academic libraries with large investments
in electronic resources. Public libraries whose operations center more on the cir-
culation of physical materials may care less about discovery products with expan-
sive coverage of scholarly articles and look more to products that provide stronger
services related to local collections and deeper levels of social engagement.
References
Breeding, Marshall. 2008. “Progress on the DLF ILS Discovery Interface API: The Berke-
ley Accord.” Information Standards Quarterly (Summer): 18–19.
Ockerbloom, John Mark (chair). 2008. “DLF ILS Discovery Interface Task Group (ILS-
DI) Technical Recommendation: An API for Interoperability between Integrated
Library Systems and External Discovery Applications.” Digital Library Federation
(Now CLIR). December 8. http://www.diglib.org/architectures/ilsdi/DLF_ILS
_Discovery_1.1.pdf.
51046_ch07_p069-086.qxd 5/19/11 12:49 PM Page 86
Related Resources
Breeding, Marshall. 2007. “Front-End Focus.” Smart Libraries Newsletter 27, no. 3
(March): 1.
———. 2007. “Next-Generation Library Catalogs.” Library Technology Reports 43, no. 4
(July/August).
———. 2007. “VuFind: A Next-Gen Catalog from Villanova.” Smart Libraries Newsletter
27, no. 9 (September): 1.
———. 2008. “Beyond the Current Generation of Next-Generation Library Interfaces:
Deeper Search.” Computers in Libraries 28, no. 5: 39–42.
———. 2008. “SirsiDynix Launches Its Faceted Search Product.” Smart Libraries Newslet-
ter 28, no. 8 (August): 3.
———. 2009. “The Advance of Computing from the Ground to the Cloud.” Computers in
Libraries 29, no. 10 (November/December): 22–26.
———. 2009. “BiblioCommons Prepares for the Next Stage of Roll-Out.” Smart Libraries
Newsletter 29, no. 8 (August): 1–4.
———. 2009. “EBSCO Sets Strategy for Discovery.” Smart Libraries Newsletter 29, no. 9
(September): 1–3.
———. 2009. “OCLC Partners with EBSCO to Expand Access to Articles in WorldCat
Local.” Smart Libraries Newsletter 29, no. 5 (May): 1–3.
———. 2009. “Summon: A New Search Service from Serials Solutions.” Smart Libraries
Newsletter 29, no. 3 (March): 1–3.
———. 2010. “Encore Synergy Launched for Article Discovery: A New Search Model.”
Information Today NewsBreak, May 3. http://newsbreaks.infotoday.com/NewsBreaks/
Encore-Synergy-Launched-for-Article-DiscoveryA-New-Search-Model-66962.asp.
86
———. 2010. “Guide to Discovery Layer Interfaces.” Library Technology Guides. Accessed
November 1. http://www.librarytechnology.org/discovery.pl.
———. 2010. Next-Gen Library Catalogs. New York: Neal-Schuman.
Dempsey, Lorcan. 2007. “Web Scale.” January 7. http://orweblog.oclc.org/archives/
001238.html.
Index Data. 2010. Company website. Accessed November 1. http://www.indexdata.com/.
NGC4LIB. 2010. Electronic discussion list. Established by Eric Lease Morgan of Notre
Dame University in June 2006. Archives. Accessed November 3. http://dewey.library
.nd.edu/mailing-lists/ngc4lib/.
Open Archives Initiative. 2010. Accessed November 1. http://www.openarchives.org/.
Serials Solutions. 2010. Summon. Accessed November 10. http://www.serials
solutions.com/summon/.
SOPAC: The Social OPAC. 2010. Website. Accessed November 10. http://
thesocialopac.net/.
Tennant, Roy. 2005. “Lipstick on a Pig.” Library Journal, April 15.
http://www.libraryjournal.com/article/CA516027.html.
VuFind. 2010. Website. Accessed November 1. http://www.vufind.org.
51046_ch08_p087-092.qxd 5/19/11 12:49 PM Page 87
Chapter 8
This chapter discusses the reasons why Koha is best suited for a public cloud
interface by dividing these reasons into three categories: deployment and admin-
istration, scalability and cost, and application exposure.
tackle the project with in-house personnel or to outsource the entire project to a
support company that can provide a complete, turnkey cloud-hosted package.
This decision will probably be driven largely by the desire and/or ability of the
library to hire the personnel necessary to provide in-house support of a Koha
installation. Larger libraries with established IT (information technology) depart-
ments will probably be in a better position to go the in-house route, although
very often one person can install and support a Koha instance. If the choice is
made to completely outsource, the library should be sure to mark up the contract
to ensure that its exact expectations are spelled out in order to avoid real or de
facto vendor lock-in. The services of a lawyer should certainly be retained if the
markup is extensive.
If the decision is to use in-house personnel, there must be the vetting and
choosing of a cloud service provider. The number of companies offering cloud
services is ever increasing. Careful consideration should be given not only to vari-
ous service plans, service level agreements (SLAs), and their associated costs, but
also to documented, past service levels (i.e., uptime, etc.). The latter is the most
important factor apart from cost. Service level history should be heavily weighted
in the selection process of a service provider, as it provides the most objective pre-
diction of future performance. Once a service provider has been selected, an
appropriate service plan must be chosen from among those available. As men-
tioned earlier, user demand and load must be carefully considered and a service
plan selected that will offer the best compromise between budget and resource
need. Plans that allow for dynamic scaling of resources should be preferred. 91
Other considerations and decisions involved in hosting Koha on a public
cloud platform include installation, support, and maintenance of Koha. These
things can be accomplished with a minimum of personnel. (This would not, of
course, include staff who perform day-to-day operations such as cataloging and
circulating.) It may be that libraries that already include an IT department may
already employ the necessary individual(s) to deploy and support a Koha cloud
installation.
A library undertaking either in-house or hosted installation should select an
individual who has a strong background in the administration of Unix-type plat-
forms. Consideration should also be given to experience in database administra-
tion as well as some level of exposure to programming with an emphasis in Perl, as
this is the language Koha is written in. This latter qualification becomes more
important if the library plans to do its own Koha development and bug fixing.
Conclusion
The time of the desktop-based library system is coming to an end. Many prod-
ucts are upgrading to cloud-based alternatives for many of the reasons discussed
here. With Koha, however, libraries get the added benefit of using an open source
application with a vibrant developer community behind it.
51046_ch08_p087-092.qxd 5/19/11 12:49 PM Page 92
References
Armbrust, Michael, Armando Fox, Rean Griffith, Anthony D. Joseph, Randy H. Katz,
Andrew Konwinski, et al. 2009. “Above the Clouds: A Berkeley View of Cloud Com-
puting.” Electrical Engineering and Computer Sciences, UC Berkeley. February 10.
http://www.eecs.berkeley.edu/Pubs/TechRpts/2009/EECS-2009-28.pdf.
djohnson. 2010. “Web-Based, SaaS, and Cloud ERP Benefits.” ERP Software at Your Ser-
vice. July 22. http://erpcloudnews.com/2010/07/Web-based-saas-and-cloud-erp-
benefits/.
Ransom, Joann, Chris Cormack, and Rosalie Blake. 2009. “How Hard Can It Be? De-
veloping in Open Source.” Code4Lib Journal, no. 7 ( June 26). http://journal
.code4lib.org/articles/1638.
92
51046_ch09_p093-108.qxd 5/19/11 12:49 PM Page 93
Chapter 9
Introduction
When most technologists discuss cloud computing, they focus on hardware or soft-
ware in the cloud. The purpose of this chapter is to discuss the advantages of data in
the cloud and data as a service, in particular, library data as a cloud-based service.
Libraries have been pooling and sharing data over a networked infrastructure since
the early 1980s. This started out as the ability to draw on a single shared and main-
tained repository for catalog records, which could be downloaded from a shared 93
repository and into a local repository. This in turn helped to build a repository of
library holdings data that libraries could use to facilitate sharing of collections. Since
that time, the types of data that libraries need to share have grown exponentially
beyond bibliographic records and library holdings. Today, libraries also need to share
electronic subscription information and information about the libraries themselves.
Unfortunately, most systems in which this library data is stored are black boxes
composed of proprietary technologies, which makes it difficult for libraries to get
any data out of most library systems. Even if a library has the local expertise to get
the data out of their systems, they then must make it available in a scalable stan-
dards-based way to others. Not only does this require expertise in metadata stan-
dards (both in and outside of libraries), web service protocols, and best practices,
it also requires a significant investment in hardware.
What libraries really need is a shared infrastructure that is capable of making
library data flexible and repurposeable. Cloud-based infrastructure lends itself
well to this scenario because cloud-based infrastructure is inherently meant to be
shareable. To allow developers to interact with them and build applications, these
systems are also inherently open and standards based.
Because of its shared metadata repository, Online Computer Library Center
(OCLC) is in a unique position to create cloud-based data services that can expose
51046_ch09_p093-108.qxd 5/19/11 12:49 PM Page 94
Leveraging OCLC Cooperative Library Data in the Cloud via Web Services
The last type of request that can be made to the WorldCat Search API is one
for holdings information. This type of request requires an identifier (OCLC
number, ISBN, and ISSN) and a geographic location be sent to the API. Geo-
graphic location can be defined based on latitude and longitude, zip code, city
and state, or IP address. If no geographic location is provided, then the API
assumes that it should retrieve holdings closest to the IP address of the application
making the request.
A library holdings request will return a list of libraries with the item requested
along with basic information about those libraries, including their name, address,
and a link to the item in that library’s catalog. Holding requests also allow users to
choose the format in which the data they are retrieving will be returned. Currently
holdings requests support output in both XML and JSON formats.
xIdentifier Services
The xIdentifiers services are a suite of services based on the WorldCat database
that associate, group, and relate items together. There are three different services
that make up the xIdentifier services: xISBN, xOCLCNum, and xISSN. The
xISBN service associates ISBNs together. As a result it can be used to show dif-
ferent editions and formats of the same work based on an ISBN that has been sub-
mitted.
xOCLCNum
xOCLCNum is a very similar service; however, rather than submitting an ISBN to 95
the service to retrieve related editions, developers can ask for related editions
based on OCLC number or LCCN (Library of Congress Classification Number).
This service is particularly helpful for materials that don’t have an ISBN. This ser-
vice also contains information on whether or not a particular publication is avail-
able as free full text from HathiTrust or Open Content Alliance. Queries to the
service can be limited to these particular full-text collections and if full texts exist
for the item requests or any other edition of the item, then a URL to that full text
will be returned.
xISSN
The xISSN web service is slightly different from the other two identifier services
because it relates together serial publications. As many librarians know, serial pub-
lications, journals in particular, can change titles over time. They can also be split
into different publications, or two publications can be merged together into a
single publication. In addition, journals can be made available in a variety of for-
mats: print, electronic, microfiche, and microfilm. The getHistory request type
can be used to find these relationships based on an ISSN. Submitting an ISSN to
this request type retrieves basic metadata about that publication including the
ISSNs for all the formats, along with any related publications such as preceding
51046_ch09_p093-108.qxd 5/19/11 12:49 PM Page 96
and succeeding titles. A getMetadata request to the web service will return whether
or not a journal is peer reviewed and a URL for the table of contents feed for that
particular publication.
WorldCat Identities
WorldCat Identities is a web service based on data in WorldCat that provides per-
sonal, corporate, and subject-based identities (writers, authors, characters, corpo-
rations, horses, ships, etc.). The service has four types of requests: direct linking
by LCCN, OpenURL, Name Search, or SRU Search. Direct linking by LCCN is
the simplest way to use WorldCat Identities. Simply send a request containing the
LCCN and a single Identity record will be returned. Identity records contain a
rich amount of metadata, including the identity’s name, authoritative name, alter-
nate names, dates of works related to identity, most popular items about and by
an identity, subject headings associated with an identity, and Wikipedia link for
the identity. Linking by OpenURL is the next simplest way to access data in
WorldCat Identities. A typical OpenURL request will contain an OCLC number
and the name of the identity. Typically OpenURL requests will result in a single
Identity record. However, if more than one Identity record matches, then the ser-
vice will return the matching records in a result Name Search response format.
Name Search is another type of request that can be made to WorldCat Identities.
This type of request takes a name parameter and searches for Identity records
matching that parameter. The service returns results in a simple XML format that
96 contains the established form of the Identity, URI, dates, date span of publica-
tions, and type of identity. Results are returned ranked based on frequency of use
of and closeness of match to name submitted.
In addition to the simpler Name Search request type, Identities also makes
available a full-fledged SRU service that allows developers to search the service
based on specific fields such as family name, first name, birth and death dates, and
LCCN. SRU Searches can be performed on the entire Identities database or on
subsets such as personal name Identities, corporate name Identities, and subject
Identities. SRU requests return results in the SRU format with embedded Identity
records. This response format contains a level of detail that is not often required by
most applications using Identities data.
Terminology Services
Terminology Services is an experimental web service from OCLC Research that
makes a variety of controlled vocabularies available via SRU. Some vocabularies
available within the service include:
• FAST
• GSA Form and Genre Terms
• Library of Congress AC Subject Headings
• Library of Congress Subject Headings
51046_ch09_p093-108.qxd 5/19/11 12:49 PM Page 97
Leveraging OCLC Cooperative Library Data in the Cloud via Web Services
WorldCat Registry
The WorldCat Registry is a database of libraries, museums, and cultural institu-
tions worldwide. Data from the WorldCat Registry is freely available for noncom-
mercial use. The database contains basic metadata about every institution,
97
including its name, address, and phone number. In addition, the Registry con-
tains more detailed information specific to libraries, such as the library catalog
software, URL for the library catalog, OpenURL resolvers, and whether or not
the library makes Wi-Fi freely available.
The WorldCat Registry can be accessed in three ways: direct linking, search,
and OpenURL resolver lookup. Developers can direct link to a registry entry
based on either Registry ID or OCLC symbol. Registry requests return informa-
tion about the library being requested in an XML format. The level of detail of
the information returned depends on whether the serviceLabel is set to content
or enhancedContent. Content returns the basic public metadata about the library,
while enhancedContent returns all the public metadata about the library.
The Registry can also be searched via SRU. The SRU interface provides a
number of indexes including institution name, library type, city, state, country, and
postal code, which developers can use to search for libraries matching the criteria
submitted. The service would then return a list of libraries matching the criteria
submitted in XML format. The XML response to this request returns basic infor-
mation about the libraries, such as library name, street address, city, state, and zip
code.
It is also possible to look up OpenURL resolvers using the Registry web ser-
vice. Resolvers can be looked up based on IP address or the “requestor” value,
51046_ch09_p093-108.qxd 5/19/11 12:49 PM Page 98
which will look in the request’s http header for IP address and submit this as the
IP address to search upon. These queries will return an XML response containing
information on the OpenURL resolvers associated with the submitted IP address.
Leveraging OCLC Cooperative Library Data in the Cloud via Web Services
called “Authors Related to Your Search” at the top of the page. Within this section
are identities that matched the search sent along with subjects related to those 99
identities. This enables users to see other possible names the author uses as well as
possible subject headings they might want to search (see Figure 9.2).
In addition to making recommendations for author searches, the VuFind Rec-
ommender module makes subject recommendations. When users perform a sub-
ject search VuFind returns a results page with the results of that search as well as a
“Subject Recommendations” section. With this section VuFind displays possible
related subjects. These subjects are drawn from the Terminology Services web
service. VuFind performs an SRU search of the Library of Congress Subject Head-
ings database within Terminology Services and returns related, broader, and nar-
rower terms in the Subject Recommendations section. By returning these subject
recommendations, VuFind is able to direct users to other possible searches that
might meet their needs and help them refine their searches.
LibX
LibX is a tool that helps libraries create a custom browser toolbar that facilitates
access to their library’s collections. LibX uses a number of OCLC web services
both to help libraries create a customized toolbar and to link users to library mate-
rials. When librarians choose to create a LibX edition they are asked to provide a
basic set of information including the URL for the library catalog and OpenURL
resolver. In the library catalog tab of the LibX edition creation screen, users can
51046_ch09_p093-108.qxd 5/19/11 12:49 PM Page 100
100
Leveraging OCLC Cooperative Library Data in the Cloud via Web Services
search the WorldCat Registry for the appropriate catalog URL for a given library
and select it (see Figure 9.3). In addition, in the OpenURL tab users can input a
hostname or IP address in order to search the Registry for the library’s OpenURL
resolver.
LibX also leverages the xISBN web service to help users find resources. If
LibX finds an ISBN within a web page it autolinks it and adds basic metadata
about the book (title, author) from the xISBN web service to a tooltip. It also helps
users search for different editions of the book with a different ISBN that the library
might hold.
Cite This
Another simple solution that several libraries have adopted based on OCLC web
services is the Cite This tool. One of the best examples is Miami University
Library, which embeds a “Cite This” button into their library discovery tool (see
Figure 9.5).
The “Cite This” button takes an OCLC number from the interface and sends
it as a Formatted Citations request to the WorldCat Search API. WorldCat Search
API then returns formatted citations for the given OCLC number, which the user
interface displays within a pop-up window. Bryn Mawr Library’s catalog has a
similar feature (http://tripod.brynmawr.edu/record=b3437799~S10). Addition-
ally, variations of this idea are being used in several libraries, including University
of Houston, Virginia Tech, Colorado School of Mines, and Universal College of
Learning. Most of these libraries link back to WorldCat.org for citation rather than
directly embedding links within the user interface, but the basic idea is the same.
Another variation is the “Cite Me” Facebook application, created by Bruce Wash-
burn of OCLC Research (http://www.oclc.org/developer/applications/citeme-
facebook-app). This application allows Facebook users to search WorldCat for
materials and get back results in various citation formats.
102
Leveraging OCLC Cooperative Library Data in the Cloud via Web Services
BookMinder Prototype
The BookMinder prototype is an Android application developed by Bruce Wash-
burn at OCLC Research. The idea behind the application was to give users a way
to scan a bar code and add that item to a personal library on their Android device.
Once an item was added to a user’s personal library the user would be able to
see libraries nearby with that item. Further information on the prototype is avail-
able on the OCLC Developer Network website (http://www.oclc.org/developer/
applications/bookminder-android-app).
iBookshelf
iBookshelf is an iPhone application quite similar to the BookMinder prototype.
Users can scan a bar code or search for a title and add it to their personal library.
Users can organize their books, track if they have loaned them to other people,
and rate books. The application allows users to find libraries near them with a
particular book based on information in the WorldCat Search API. Information
about that library and its location is then available to the user (see Figure 9.7).
104
Leveraging OCLC Cooperative Library Data in the Cloud via Web Services
CampusBooks
CampusBooks is an application for both iPhone and Android that allows users to
scan a bar code or enter a title, author, keyword, or ISBN in order to search for
books online, compare prices, and find books in local libraries through the World-
Cat Search API. Once users finds the item they want they can contact the local
library through information provided by the WorldCat Registry API, such as
address, phone, and URL (see Figure 9.8).
105
106
Pic2Shop has also adopted the WorldCat Search API in order to connect users
with libraries. This application is downloadable globally and shows results for
libraries worldwide (see Figure 9.9).
Conclusion
Data services in the cloud allow library data to be repurposed and leveraged in
new ways. For example, OCLC’s cloud-based data services provide multiple ser-
vices based on the data in WorldCat. WorldCat Search API, WorldCat Identities,
and the xIdentifier services are all fundamentally built on WorldCat data. How-
ever, each service represents an optimized and enhanced view of the WorldCat
data for a particular purpose.
Furthermore, as is evident from the variety of applications discussed in this
chapter, cooperative data in the cloud offers several distinct advantages. It creates
a shared infrastructure to distribute library data to members of the cooperative,
51046_ch09_p093-108.qxd 5/19/11 12:49 PM Page 107
Leveraging OCLC Cooperative Library Data in the Cloud via Web Services
Resources
Developer Network Application Gallery. http://www.oclc.org/developer/applications.
Terminology Services Documentation. http://www.oclc.org/developer/documentation/
terminology-services/using-api.
WorldCat Identities Documentation. http://www.oclc.org/developer/documentation/
worldcat-identities/using-api.
WorldCat Registry Documentation. http://www.oclc.org/developer/documentation/
worldcat-registry/using-api.
WorldCat Search API Documentation. http://www.oclc.org/developer/documentation/
worldcat-search-api/using-api.
xISBN Documentation. http://www.oclc.org/developer/documentation/xisbn/using-api.
xISSN Documentation. http://www.oclc.org/developer/documentation/xissn/using-api.
xOCLCNum Documentation. http://www.oclc.org/developer/documentation/xoclcnum/
using-api.
51046_ch09_p093-108.qxd 5/19/11 12:49 PM Page 108
51046_ch10_p109-122.qxd 5/19/11 12:49 PM Page 109
Chapter 10
Building Push-Button
Repositories in the Cloud
with DSpace and Amazon
Web Services
John Davison
Introduction
The Digital Resource Commons (DRC) is a multi-institution academic repository
for archiving, discovering, and sharing unique academic materials produced by
the University System of Ohio and Ohio’s private colleges. The DRC uses DSpace
open source software to store an ever-expanding set of digital collections. At the 109
present time, the DRC hosts more than half a million items from 39 institutions
and serves more than three million page views annually to visitors from all over
the world. This chapter focuses on the experiences of the DRC in the Amazon
cloud from June 2009 to December 2010.
The DRC development team has created a freely available, preconfigured ver-
sion of their repository software that will allow other organizations to replicate the
DRC’s cloud-based procedures for building digital collections and take advantage
of Amazon Web Services (AWS) to reduce the technical barriers to digital object
storage and preservation. Hosting these repositories in the cloud environment
removes the need for expensive hardware investments and simplifies site manage-
ment, allowing more organizations to preserve their unique digital heritage. This
chapter will outline some of the uses and advantages of cloud-based repository
services and will describe how others can use the DRC software to quickly create
digital collections.
continue building critical mass for the project while financial and technical road-
blocks prevented its parent organization from allocating additional computing
resources. The answer was to use the cloud to divorce application architecture
from underlying hardware and build a repository framework that masked the ser-
vice’s complexity from the end user. At the same time, virtualization allowed the
DRC team to create a process with push-button simplicity. Relying on the Amazon
cloud to provide the hardware on a pay-as-you-go basis, the DRC was able to
quickly create new institutional repositories for eight members: Kent State Uni-
versity, Kenyon College, Xavier University, Ohio Wesleyan University, Ohio Uni-
versity, the University of Cincinnati, and the two education-related groups The
Digital Archive of Literacy Narratives and the College & University Disability
Access Collaborative. The majority of member organizations are part of the Ohio-
LINK library consortium, but the DRC is broadly tasked to include nonmember
educational and historical communities such as The Archives of the History of
American Psychology and the Rutherford B. Hayes Presidential Center.
One of the key indicators of success is the inability to distinguish cloud-based
instances from internally hosted ones. At the same time, each repository is seam-
lessly integrated with its parent institution’s branding, and federated searching is
possible across all repositories. DSpace instances can be brought online for new
members in less than 10 minutes, and they can begin submitting content immedi-
ately.
This chapter serves as an introduction to the DRC’s cloud-based repository
110 development, providing details for creating an AWS account and step-by-step
instructions to launch instances of the DRC software. Additional administrative
steps are briefly outlined to assist the reader in moving toward a production-ready
repository. Possible future options for cloud computing, open source repository
software, and the DRC project are also discussed.
Technology Policy has named the DRC a 21st Century Cutting-Edge Technol-
ogy Service for 2011.
form an administrative function, they provide the matching private key. Security
key pairs can be one of the more confusing aspects of AWS for new users, but they
are relatively easy to create in the Management Console. For this example, use the
name “EC2 Account.” Under the Networking & Security section, click on the Key 113
Pairs link on the left-hand side. Generating the RSA key pair will download a file
with a .PEM (privacy enhanced mail) extension (see Figure 10.3). Because this
file provides the needed credentials to access all aspects of instance management,
114 take care to protect it.
and the Instance Type. By launching multiple instances of an application in dif- 115
ferent availability zones, it is possible to insulate operations from failure in a single
location. For this example, select No Preference. Choose the smallest instance
type, the Micro with 613 MB of RAM. Make sure the “Launch Instances” radio
button is selected and press Continue (Figure 10.5).
The next screen of advanced instance options can be left at default values for
Kernel, RAM disk, and CloudWatch. Press Continue. The wizard will now ask
for an instance name. Type “DRC Repository” in the Value field and continue.
Verify the EC2 Account key pair generated earlier is being used. There is no need
to change the default security group for this exercise. Review the options screen,
and launch the new instance (see Figure 10.6).
116
Figure 10.6. Review Options and Launch
51046_ch10_p109-122.qxd 5/19/11 12:49 PM Page 117
camps. Efforts are already underway to build a DSpace release that runs on top
of a Fedora repository. This, along with the DuraCloud-hosted service, is widely
seen as a major step toward building an entirely modular repository platform able
to more easily leverage cloud computing options (Donohue, 2010a).
rapid development of technical proficiency that can be shared throughout the state
and with the DSpace community at large; and a compelling model for more inde-
pendent institutional control and access to site administration. Interested parties
can follow the work of the DRC Development Team at the Files in This Item blog
(http://fiti.ohiolink.edu).
References
Cunningham, Lewis. 2009. “Using and Managing AWS—Part 3: AWS Security.” May 17.
http://clouddb.info/2009/05/17/using-and-managing-aws-part-3-aws-security/.
Donohue, Tim. 2010a. “DuraSpace Recommends DSpace & Fedora Integration Initia-
tive.” DuraSpace (blog), July 8. http://expertvoices.nsdl.org/duraspace/2010/07/08/
duraspace-recommends-dspace-fedora-integration-initiative/.
———. 2010b. “Making DSpace 1.5.x Your Own.” Accessed October 5. http://www
.slideshare.net/tdonohue/making-dspace-15-your-own-customizations-via-overlays.
Fedora Reference Repository. 2010. DuraSpace (website). Accessed October 10.
http://testdrive.fedora-commons.org/ref-repo.php.
Fingerhut, Eric. 2010. “Strategic Plan for Higher Education 2008–2017.” Accessed Octo-
ber 2. http://www.uso.edu/strategicplan/.
MacVittie, Lori. 2010. “Infrastructure 2.0 + Cloud + IT as a Service = An Architectural
Parfait.” Two Different Socks (blog), September 15. http://devcentral.f5.com/
weblogs/macvittie/archive/2010/09/15/infrastructure-2-enables-cloud-enables-it-as-
a-service.aspx.
Maslov, Alexey, James Creel, Adam Mikeal, Scott Phillips, John Leggett, and Mark McFar-
land. 2010. “Adding OAI-ORE Support to Repository Platforms.” Journal of 121
Digital Information 11, no. 1. http://journals.tdl.org/jodi/article/view/749.
Morris, C.M. 2008. “Can Anything as Ephemeral as a ‘Cloud’ Provide Permanent Stor-
age?” Content Sausage (blog), November 12. http://blogs.nature.com/uc0af3996/
2008/11/12/can-anything-as-ephemeral-as-a-cloud-provide-permanent-storage.
Phillips, Scott, Cody Green, Alexey Maslov, Adam Mikeal, and John Leggett. 2007. “Man-
akin: A New Face for DSpace.” D-Lib Magazine 13, no. 11/12. http://www.dlib.org/
dlib/november07/phillips/11phillips.html.
51046_ch10_p109-122.qxd 5/19/11 12:49 PM Page 122
51046_ch11_p123-132.qxd 5/19/11 2:18 PM Page 123
Chapter 11
Untethering Considerations:
Selecting a Cloud-Based
Data Access and
File-Sharing Solution
Heidi M. Nickisch Duggan
and Michelle Frisque
Introduction
Many librarians need to access files and data stored on their office hard drives
from remote locations, even when they don’t have access to their campus or library 123
networks. In some cases files need to be accessible regardless of computer operat-
ing system. Staff librarians often also need to share files of all sizes with one or
more colleagues and collaborators, not all of whom are part of the same institu-
tion. A number of excellent no- to low-cost cloud-based solutions are on the
market that provide library staff with an easy way to securely access, share, and
work on files from anywhere to ensure that the most recent version of each file is
automatically available on any other computer that they use.
Variables to Consider
There are a number of variables to consider while selecting and implementing a
cloud-based solution for file synchronization. Factors such as what data staff need
to have remote access to and how much space is required, who needs access,
whether the files require a private or a collaborative space or both, and if the prod-
uct is priced affordably will guide the product selection itself. The library’s com-
puter support department will have concerns and questions regarding product
administration and management, installation and user training, platform support,
and restoration. The various IT units of the broader organization are sure to have
requirements regarding encryption and other technical and legal compliance
issues. Aspects librarians should consider when evaluating and selecting a cloud-
51046_ch11_p123-132.qxd 5/19/11 2:18 PM Page 124
based file storing/sharing solution include web-based, client-side, and hybrid solu-
tions; administration and management of the system; ease of use; compliance;
encryption; pricing structure; data restoration options; file type support; vendor
support; and user recommendation.
syncing so that once users save a file in a specific location on their com-
puter it is automatically synced to all of their other accounts. The web-
based solution usually requires that users remember to upload the most
recent version of the document.
• Synchronization is a useful feature that allows seamless access to the
most current version of a document regardless of the computer on which
it was last edited. If the solution selected has an option to sync files auto-
matically, does the solution sync the entire file each time it is modified,
or does it sync only the changes that are made to the file? Syncing only
the changes to a file uses less bandwidth than saving the entire file each
time a change is made. Can the users manually customize the bandwidth
limits if they determine that the default settings are slowing down their
Internet connection? If the users edit a file when the computer is not
attached to the Internet, do they have to manually sync the files, or will
the solution sync the changes automatically once an Internet connection
is reestablished?
Ease of Use
A primary consideration for any organization deploying a new product is how
easy it is for end users to learn and use as well as how easy it is for system admin-
istrators to manage. In the case of file access and synchronization, evaluate how
straightforward it is to save, synchronize, organize, access, and share files. Doing
some preliminary testing with the user base will let the computer support staff
know whether users find the interface intuitive or whether they will need to pro-
vide training and a more gradual rollout of the product. The more difficult a prod-
uct is to use the less likely staff will adopt the service. If the product cannot easily
be installed, customized by the user, and implemented or it requires a system
51046_ch11_p123-132.qxd 5/19/11 2:18 PM Page 126
administrator to set up and maintain, then computer support personnel will need
to be available and prepared to support the end users.
Encryption
When storing data using a cloud-type service, ensure that the data is encrypted
when it is synced and stored on the vendor’s servers. Most vendors automatically
encrypt the data that is synced and stored on their servers, but some organizations
126 may require an additional level of encryption using specific third-party encryp-
tion software. If the library’s parent organization requires the use of encryption
software like PGP Data Protection or Symantec Endpoint Encryption, make sure
the file sharing solution selected supports third-party encryption.
Pricing Structure
Pricing structures vary by product, and it is often simpler and more cost effective
to purchase products that have an organizational pricing model rather than paying
for each individual account. Consider whether the product is available as a one-
time purchase, is subscription based, or is part of a larger package. The fee struc-
ture may also vary by number of users, storage space required, or other criteria.
Many vendors will accept institutional purchase orders, but some will require a
personal or company credit card.
• How easy is it to recover a file that has been deleted? For how long are
deleted files stored in the system?
• Is it possible to restore previous versions of a file? How long are previ-
ous versions of the document available? Some file sharing solutions keep
every saved change of the file for up to 30 days. Some keep only a fixed
number of saved changes of the document, and yet others keep the
changes indefinitely, sometimes for an additional fee.
• Does file restoration require an administrator’s support, or can the user
complete it with ease?
• Will the system allow the restoration of more than one version of the
same file?
• Does the product provide the ability to customize any of these settings
for each user, or is that accomplished at the group level? If the library
has an organization subscription, can these settings be customized in
the administrative tool?
• If the organization has an institutional account, can the administrator
restore deleted files as well as restore previous versions of existing files,
or can this be done only at the individual account level?
Vendor Support
Even with solutions that are easy to use, there are times that the administrator or
individual users may need additional support in troubleshooting issues with the
cloud-based file sharing solution. Things to consider when thinking about future
support issues include:
• Does the vendor have online documentation? Before selecting a solu-
tion, read some of the documentation to make sure it is easy to under-
stand.
51046_ch11_p123-132.qxd 5/19/11 2:18 PM Page 128
• Does the vendor have an active user group or forum that staff or the
system administrator can consult to get help from other users?
• Can the library’s support staff contact the vendor directly for support? Is
the vendor’s contact information easy to locate on its website? What is
the library’s support staff ’s preferred communication channel and does
the vendor offer that option (i.e., e-mail, chat, phone)? Is there an addi-
tional fee related to any of these services? Does the vendor offer fee-
based support packages that will provide the level of service you want?
Is the price reasonable?
Box.net
Box.net (http://www.box.net) is a cloud-based file sharing tool libraries should
investigate if the primary need is to support and maintain a large number of user
accounts. Of all of the options explored in this chapter, this one is the most robust.
Box.net has several different plans for organizations to consider. The Personal
option is for individuals and is free. It includes 5 GB of web storage and any single
file cannot be larger than 25 MB, although additional storage can be purchased.
The Personal account option does not support syncing; instead, files need to be
uploaded to the Box.net website. Files and folders can be shared with others via a
web link. To make collaboration easier users can create wiki-style documents that
allow multiple people to edit and comment on documents directly online.
The Box.net Business and Enterprise solutions include all of the features avail-
able to the Personal account as well as many others. The greatest benefit of the
Business and Enterprise solutions is that they support syncing files from the desk-
51046_ch11_p123-132.qxd 5/19/11 2:18 PM Page 129
top to Box.net. Users also have the option to access files via the Box.net website.
The Business and Enterprise solutions also allow users to track file version. Other
features available to both the Business and the Enterprise solution include full-
text search, statistics, administrative console, and Google Apps integrating. The
Business option begins at a 500 GB space limit for a monthly per-user subscrip-
tion and includes a 2 GB file size restriction.
The Box.net Enterprise solution includes all of the features available to the
Personal and Business account. Additional features that are available only to Enter-
prise customers include custom branding, role-based access control, additional
level of data encryption, bulk account creation and editing, an option for Active
Directory/LDAP integration for account authentication, and dedicated support
from Box.net.
Dropbox
Dropbox (http://www.dropbox.com) is a cloud-based syncing and backup prod-
uct for any file size or type. It is a hybrid solution with both client and web access.
Dropbox is free for up to 2 GB of storage, with up to 100 GB of storage available
for a fee. Users can also increase their free storage by inviting friends to join Drop-
box or by purchasing additional space, available on an annual basis. Dropbox for
Teams allows shared storage for a workgroup and has an administrative console for
centralized management of the accounts.
Installing and configuring Dropbox is relatively simple; users are generally
able to implement Dropbox without extensive IT support. To get started each 129
Dropbox user needs to create an account on the Dropbox website. Then they
install the Dropbox client software on all of the computers they want to access
their Dropbox files from. As part of the software installation process, Dropbox
creates a “My Dropbox” folder (by default placed in the My Documents folder on
a Windows machine and the Finder window on the Mac). All files and folders
saved in the Dropbox folder are synced in the cloud, and then Dropbox syn-
chronizes those files to any other computers (Windows, Mac, or Linux) or mobile
device (iPhone, iPad, Android, and Blackberry) where Dropbox is installed with
that account. The syncing is automatic, occurring on-the-fly whenever the com-
puters or devices are connected to the Internet, which means users always work
on the most recent version of the file. Dropbox also allows users to work on their
files in their Dropbox account even when they are not online. Any changes made
to a file in Dropbox sync to the Dropbox account as soon as the computer is
reconnected to the Internet.
Dropbox is also accessible from a web browser via a computer or mobile
device—even if the Dropbox client is not installed—so files are always at hand.
The Dropbox website allows users to access files when they are not at their own
computers or if they do not have the client installed. By logging in to the website,
users can see all of their files and folders in the same hierarchy used on their hard
51046_ch11_p123-132.qxd 5/19/11 2:18 PM Page 130
drives. However, to edit files, users need to download them to the computer they
are working on and then save them when they are finished and upload them to
the Dropbox website. For more information about a library’s use of Dropbox, see
Chapter 15 of this book.
Google Docs
Google Docs (http://docs.google.com) is another option that some libraries have
explored. Previously, Google Docs supported only a limited number of file types,
and the files had to be converted to an available Google Docs format (i.e., Docu-
ment, Presentation, Spreadsheet, Form, and Drawing). Early in 2010, Google
started allowing users to store up to 1 GB worth of files that are not converted to a
Google Docs format for free and anything above that for a fee.
All files must be uploaded via a Google webpage to be stored in Google Docs.
A few browsers, including newer versions of Safari, Chrome, and Firefox, support
the ability to drag files from the user’s desktop. When users upload files to Google
Docs they can choose to retain the file type or they can convert the uploaded file
into the corresponding Google Docs format. All documents that are converted to a
Google Docs format can be edited online; however, some file types cannot be con-
verted to a Google Docs format. Examples of files that can be converted to a
Google Doc format include, but are not limited to, .doc, .docx, .odt., sxw, .rtf, .txt,
.ppt, .pps, .xls, .xlsx, .csv, and .wmf. When users choose to retain the file format
that is uploaded and do not convert it to a Google Doc format, they will need to
130 either download the file to edit it or convert it to a Google Doc format to edit the
file online.
Files uploaded to Google Docs can be organized in folders created by the user.
Files that are uploaded to Google Docs use the same permissions system that is
used for other native Google Docs documents. The permissions can be set at an
individual file level or at the folder level. There are three ways to share files with
others. The first option is the Public on the Web option, which allows anyone on
the Internet to view the document without signing in. The Anyone with a Link
option allows anyone with a special link to access the file. The Private setting is the
most restrictive in that it allows access to a file only to individuals who have been
explicitly granted permission. All users must have a Google account to log into
these files. While the user can share an uploaded file, only files that are converted
to a Google Docs format can be edited online.
SkyDrive, some file types have additional features. For instance, Microsoft Office
files can be created, edited, and stored within Windows Live SkyDrive using the
Office Web App suite available online within the Windows Live Suite. Images and
video can also be edited using online software within the Windows Live suite.
Windows Live SkyDrive includes 25 GB of free storage. The Windows Live
SkyDrive software allows users to synchronize and access files across multiple
computers. Users can choose to sync individuals’ files or specific folders on their
desktop. Desktop syncing requires a client to be installed on the computer. It cur-
rently supports both Windows and Mac operating systems, and files can also be
accessed online via the Windows Live service. Any files stored online in Windows
Live SkyDrive can be shared with individuals and groups or may be kept private.
All files can also be accessed via the Windows Live website. Windows Live Sky-
Drive does not currently have the option to manage group accounts.
Conclusion
A cloud-based file sharing solution can be an easy to implement, no- to low-cost
solution, providing library staff with an easy way to securely access files from any-
where and the ability to seamlessly share files with colleagues and collaborators.
To select the most appropriate cloud-based file sharing solution, organizations
should assess their users needs to make sure that they select the most appropriate
solution for their organization.
Consider what data staff need to have remote access to and how much space
that data requires, whether the files require a private or a collaborative space or 131
both, and whether the product is priced affordably. The library’s computer sup-
port department should also consider the impact of the administration and man-
agement of the selected solution, installation and user training, platform support,
and data restoration. The various IT units of the broader organization might need
to be consulted regarding the organizational requirements for encryption and
other technical and legal compliance issues. Fully understanding the library’s
needs will help each library evaluate the various cloud-based file sharing solutions
available so that the library selects the option that not only fits its needs of today
but seamlessly supports its needs for the future.
51046_ch11_p123-132.qxd 5/19/11 2:18 PM Page 132
51046_ch12_p133-140.qxd 5/19/11 2:19 PM Page 133
Chapter 12
SharePoint Strategies
for Establishing a Powerful
Library Intranet
Jennifer Diffin and Dennis Nangle
Introduction
In the world of cloud computing, libraries have many options for setting up a
remotely accessible intranet, which can become a collaborative space that is flexi-
ble to a library’s organizational needs. Many factors should be taken into consid-
eration, including, but not limited to, security, accessibility, and features. Microsoft
SharePoint is one option to be considered. Regardless of whether the servers being
used to store the information are in house or outsourced, users can collaborate 133
with colleagues regardless of physical location through their web browser. Share-
Point is available as part of the Microsoft Core Access License (CAL) Suite for
institutions (Gilbert, 2010). However, as Gartner analyst Mark Gilbert (2010) has
mentioned, one SharePoint flaw is the lack of documentation. This chapter will
outline strategies that libraries can utilize to develop an efficient and effective
intranet that combines social networking tools like blogs, wikis, and discussion
boards with dynamic knowledge management functionality.
Defining SharePoint
SharePoint can be defined overall as a proprietary collaborative workspace, owned
by Microsoft and manifesting itself differently depending on the type of license
purchased and implementation method. A couple of terms that often surface in
the SharePoint environment are “Sites” and “Pages.” While any user with a basic
familiarity of the Internet can recognize these terms, it is crucial to understand
their key distinctions within SharePoint. A Site in SharePoint carries with it much
of the expected functions of a website in the traditional sense: it is the main loca-
tion under which several Pages of a similar category would be placed. However,
SharePoint’s various Site categories, which a user must select upon creating a
SharePoint Site, significantly affect functionality moving forward. In SharePoint
2010, there are five categories of Sites available for hosting and managing infor-
mation: Team Sites, Document Workspaces, Group Work Sites, Meeting Work-
spaces, and Blog Sites (Bates and Smith, 2010). Categories may vary based on the
version of SharePoint and the individual installation at an institution. Within a
Site, there can be one or more components (Libraries, Lists, and Pages) available
to use to add content to the Site (Bates and Smith, 2010).
Most SharePoint Sites begin with a blank Page. Using the Internet terminology,
this can be likened to a homepage. From here various types of Pages and other
134 content can be created and added to the Site. The Page types available are Page
(SharePoint 2010 only), Publishing, and Web Part. A SharePoint List is essen-
tially a simple database or spreadsheet that contains information or data that is
collected in rows. Each row is called a “list item” (Sagi, 2010). Many of the fea-
tures discussed in the chapter are based on the SharePoint List functionality:
Tasks, Announcements, Discussion Boards, Calendar, etc. These Lists are very
flexible and allow staff to customize them to best suit the needs for which they are
being used. A SharePoint Library is a particular kind of List where each list item is
one file (image, form, document, video, etc., depending on type of SharePoint
Library). Sagi (2010) provides an example of an institution that could use Share-
Point as an intranet, with the home Site housing relevant information for organi-
zational staff as a whole. Then, Sagi (2010: 7) explains further, it may link to “a
subsite called Human Resources that stores forms such as travel requests, expense
claims, and other forms. . . . The two sites may share some attributes . . . but they
have separate contents—for example, different pages, libraries, and lists.”
Implementation Strategy
Establishing the intended purpose of SharePoint before implementation is
extremely important. To help ensure that SharePoint is high on the priority list, the
implementation project’s goals and objectives should be clearly defined; agreed
51046_ch12_p133-140.qxd 5/19/11 2:19 PM Page 135
on by the key stakeholders; and tied to the mission and goals of the institution
and/or group that will be implementing SharePoint.
As implementation of SharePoint progresses, understanding the specific objec-
tives that are trying to be achieved will have an impact on which features are imple-
mented. Each objective may have several corresponding SharePoint features that
may be useful, so it is important to communicate with staff how a particular feature
is expected to be used. To reduce possible resistance among staff to adopting
SharePoint, be sure to have a clear communication plan and offer training sessions
that explain how using particular SharePoint features will improve on existing
workflows and systems (Jamison, Hanley, and Cardarelli, 2011).
SharePoint Features
Wiki
Regardless of context (special, academic, public, etc.), most libraries face the issue
of how to consolidate documentation. Is institutional knowledge evenly distrib-
uted and accessible with paper manuals or sporadic sticky notes in cubicles? By
51046_ch12_p133-140.qxd 5/19/11 2:19 PM Page 136
fully exploiting the Wiki feature in SharePoint, libraries can create and easily main-
tain a consolidated knowledge base. SharePoint allows for easy editing and updat-
ing through the WYSIWYG (What-You-See-Is-What-You-Get) editor.
Collaboration can be encouraged with the knowledge that any errors can quickly
and easily be undone via the version history feature. Each time a Page is saved
with new changes, SharePoint creates a new version of the Page in the background
and keeps track of what changes were made. When viewing a previous version, all
changes are marked in red for easy identification. Staff can be automatically alerted
to changes through e-mail alerts. Access to certain Pages, such as those that con-
tain passwords and private user data, can be controlled by permission levels. Sen-
sitive information can still be added to the Wiki with the knowledge that it can be
securely locked down.
With the SharePoint Wiki, text can be written concisely when libraries utilize
visual aids. Most training manuals regarding library procedures function as a
quick-reference tool as well as a supplement to the training process. Therefore,
by frequently inserting screenshots libraries can replace exhaustive body text and
increase the usability of their manual. Libraries can take advantage of SharePoint’s
Photo Library feature to enhance the visual component of their Wiki manuals.
The role of the Photo Library function as it relates to efficient Wiki updating and
insertion of screenshots will be discussed in greater detail later in “Other Share-
Point Libraries.”
136 Discussion Board
The proliferation of interdepartmental communication tools has certainly been
beneficial to efficiency among coworkers. However, these diverse channels create
“leaks” in important information; two employees discuss a procedural issue over
IM but do not think to transfer the communication in a group e-mail or blog post.
SharePoint’s Discussion Board allows for a complete record of valuable depart-
mental interactions that do not occur in a vacuum. Discussions are consolidated in
one location in a threaded message format and can be stored for as long as they are
needed. At the University of Maryland University College Library, the Document
Management Team “uses the discussion board to build consensus on issues that
require group decision making. Team members can respond to discussions while
working on other tasks at their desks, eliminating the need for time-consuming
meetings” (Diffin et al., 2010: 233). Staff members are notified of new discussions
and responses through e-mail alerts.
Calendar
Utilizing the Calendar feature on a library’s remotely accessible SharePoint Site
contributes to open lines of communication. Library staff can update their sched-
ules on-the-fly to account for absences and sudden schedule shifts, which imme-
diately informs relevant departments through the e-mail alerts feature. Another
51046_ch12_p133-140.qxd 5/19/11 2:19 PM Page 137
use for the Calendar is tracking “project milestones and deadlines along with the
inevitable team meetings and other more social get-togethers that are scheduled
during the period of the collaboration” (Harvey, 2009: 311). Depending on the
versions of the software that are being used, the SharePoint Calendar can be inte-
grated with Microsoft’s Outlook e-mail software. In remote situations, updating
the SharePoint Calendar is often easier than updating the Outlook calendar via
the web client.
Announcements
An effective replacement for e-mail blasts, the Announcements feature ensures
immediate mass notification to an entire team, while also automatically preserv-
ing the information in a location more authoritative and secure than an individ-
ual’s e-mail inbox. For a project, Announcements provide a quick and easy way
to alert team members of changes or updates, as well as a way to provide encour-
agement and support (Harvey, 2009). Staff can stay up-to-date on their depart-
ment’s happenings through various channels, including customizable e-mail alerts
and RSS feeds.
Document Library
The issue of document storage is a common struggle within library departments.
While emerging solutions have initially been promising (e.g., shared networked
drives and local intranets), these stand-alone systems remain static and are thus not
conducive to collaboration. While the SharePoint Document Library seems like 137
simply a storehouse of pertinent files, the cloud environment in which these files are
stored allow for unprecedented collaborative efficiency and flexibility. Users can
upload documents directly in SharePoint or through the use of a mapped network
drive. Documents can also be “checked out,” which officially locks them until they
are “checked in” again. Workflows can be created to set up routing for document
review. Each document has a unique and permanent URL, which makes sharing
and linking to the document very simple. Like the Wiki, it has version history and
granular permission settings. This ensures secure data storage and access in the
cloud while allowing and encouraging collaboration. With version history enabled,
restoring a previous version is as easy as a few clicks of the mouse. And the granular
permission settings allow management to lock down documents to only staff who
should have access to read or edit a particular document. There is also a Recycle
Bin where deleted documents can be recovered.
Similar to implementing a Wiki in SharePoint, the Document Library’s use
should be planned before moving any documents into it. A few things to consider
are naming conventions, retention policy, and version history control. At the Uni-
versity of Maryland University College Library, the Document Management Team
originally simply moved all of their documents from a shared network drive into
the SharePoint Document Library. They quickly realized the Document Library
51046_ch12_p133-140.qxd 5/19/11 2:19 PM Page 138
would become just as messy as the old shared network drive if they did not step
back and plan a hierarchy and structure for the documents. Many documents
ended up becoming Wiki Pages based on this planning and restructuring (Diffin,
Chirombo, and Nangle, 2010). Without these strategies, even with its high-level
functionality, a library’s SharePoint Document Library could become as ineffective
as the systems that came before it.
Task List
Task-tracking systems tend to vary greatly, from formalized task lists in Microsoft
Outlook and Project to a carefully arranged physical stack of sticky notes. While
these approaches are effective, and certainly prove to be acceptable methods of
managing tasks, SharePoint’s Tasks feature brings a team element to project man-
agement. Library organizations are beginning to embrace a “clan culture . . .
[which] provide[s] a more supportive environment for innovation and risk-
taking” (Kaarst-Brown et. al., 2004: 42). This particular culture survives by indi-
viduals regularly collaborating to get work done. With this emerging team
approach, it is imperative that project deadlines and details be uniformly shared
with the group.
Teams can use SharePoint Tasks to efficiently keep team members aware of
project progress. If a team has a SharePoint group formed, individuals can post a
Task that is assigned to the team in general, with all members receiving alerts to
51046_ch12_p133-140.qxd 5/19/11 2:19 PM Page 139
various Task creations and progress updates. Having the ability to assign Tasks
to groups of individuals can allow team members to use the Tasks List as a kind of
asynchronous brainstorming room for projects.
Blog
Similar to the Tasks feature, employing the Blog function in SharePoint has the
potential to serve as an effective multi-team sounding board, while still maintaining
the capability to speak to more granular sub-teams and committees. People can
create individual Blogs that they can use to keep others informed on specific topics
or the progress of projects. Team Blogs are a great way for team members to dis-
cuss projects and post project updates. A more general library Blog can also be
created to allow staff throughout the organization to contribute and provide feed-
back on topics of discussion. Each Blog entry can be assigned a category that can
help staff find discussions that are of interest to them. Permission levels can be set
to allow entries and/or comments by specific people or groups. An approval
process can also be implemented that would ensure content is reviewed before
being posted.
Conclusion
Microsoft SharePoint is a powerful collaborative tool that is certainly difficult to
encapsulate in one chapter. However, the intended outcome is not to equip the
reader with a comprehensive skill set in the technical features of SharePoint.
139
Instead, use the strategies described to get the innovative wheels spinning based
on the context of your particular library’s intranet needs. SharePoint’s open-
ended approach of providing several customizable options to the user can be
overwhelming. However, take advantage of this customization to tailor features
that will speak directly to your organizational workflow. Refer to the References
cited to learn more about carrying out specific technical functions within Share-
Point.
While these features and functions work effectively on their own, SharePoint’s
greatest and most long-lasting strength is when it is used as a comprehensive solu-
tion. Simply using SharePoint as a Wiki or a documentation platform may overwhelm
and complicate matters for staff members, as they may see SharePoint as just another
website or login they have to remember. Using every feature within SharePoint to
create a complete intranet solution will exponentially increase its success within an
organization. Even if buy-in is not evident (or even possible) in every library depart-
ment, start small and allow these successes to influence wider adoption.
References
Bates, Seth, and Tony Smith. 2010. SharePoint 2010 User’s Guide: Learning Microsoft’s
Collaboration and Productivity Platform. Berkeley: Apress. http://common
.books24x7.com/book/id_35235/book.asp.
51046_ch12_p133-140.qxd 5/19/11 2:19 PM Page 140
Diffin, Jennifer, Fanuel Chirombo, and Dennis Nangle. 2010. “Cloud Collaboration: Using
Microsoft SharePoint as a Tool to Enhance Access Services.” Journal of Library
Administration 50, no. 5: 570–580. http://www.informaworld.com/10.1080/
01930826.2010.488619.
Diffin, Jennifer, Fanuel Chirombo, Dennis Nangle, and Mark de Jong. 2010. “A Point to
Share: Streamlining Access Services Workflow through Online Collaboration, Com-
munication and Storage with Microsoft SharePoint.” Journal of Web Librarianship 4,
no. 2: 225–237. http://www.informaworld.com/10.1080/19322909.2010.501278.
Gilbert, Mark. 2010. Making SharePoint Deliver (webinar). Gartner. http://mfile
.akamai.com/23543/wmv/citrixvar.download.akamai.com/23543/www/900/613/5585
308717325900613/1-5585308717325900613-12ac8ced095.asx.
Harvey, Greg. 2009. SharePoint 2007 Collaboration for Dummies. Hoboken: Wiley.
http://proquestcombo.safaribooksonline.com/9780470413425.
Jamison, Scott, Susan Hanley, and Mauro Cardarelli. 2011. Essential SharePoint 2010:
Overview, Governance, and Planning. Upper Saddle River, NJ: Addison-Wesley.
http://proquestcombo.safaribooksonline.com/9780321700827.
Kaarst-Brown, Michelle, Scott Nicholson, Gisela von Dran, and Jeffrey Stanton. 2004.
“Organizational Cultures of Libraries as a Strategic Resource.” Library Trends 53, no.
1: 33–53. http://hdl.handle.net/2142/1722.
McDonald, Kevin. 2010. Above the Clouds: Managing Risk in the World of Cloud
Computing. Cambridgeshire: IT Governance Publishing. http://proquestcombo
.safaribooksonline.com/9781849280310.
Sagi, Ishai. 2010. SharePoint 2010 How-To. Indianapolis: Sams. http://
proquestcombo.safaribooksonline.com/9780132487481.
140 Salido, Javier, and Doug Cavit. 2010. A Guide to Data Governance for Privacy, Confiden-
tiality, and Compliance. Part 5: Moving to Cloud Computing. Redmond, WA:
Microsoft. http://go.microsoft.com/?linkid=9740995.
51046_ch13_p141-152.qxd 5/19/11 2:19 PM Page 141
Part III
CASE STUDIES
51046_ch13_p141-152.qxd 5/19/11 2:19 PM Page 142
51046_ch13_p141-152.qxd 5/19/11 2:19 PM Page 143
Chapter 13
Introduction
The idea of digital preservation is nothing new. What is new is the need for
libraries to preserve digital objects for an indeterminate period of time. Systems
backups were always required. These were focused, however, on restoring imme-
diate access after something went wrong. Digital preservation must look at longer 143
goals. According to William Dougherty (2009: 599-602) in “Preservation of Dig-
ital Assets: One Approach,” “Britain’s National Archive holds the equivalent of
580,000 encyclopedias of information in file formats that are no longer commer-
cially available.” Digital preservation is more than just bringing a system back up, it
is planning for a time when the very technologies currently in use will be obsolete
and some future user with unknown technology must try and retrieve that data.
Many large libraries and institutions are working on these issues at a monumental
level. Projects such as Portico (http://www.ithaka.org/portico) seek to archive mas-
sive collections of e-journals. For preserving individual digital objects there are
commercial products such as OCLC’s Digital Archive (http://www.oclc.org/digi-
talarchive/). Other libraries have built custom solutions, such as the Florida Digi-
tal Archive (Caplan, 2010). This works well if you have the technical expertise in
house to create custom solutions, but for a library like ours, another way was
needed. To do this we created a system that allows us to archive digital objects
securely both locally and to the cloud. We also designed a way to make sure that
those objects could be found and retrieved in the same condition in which they
were deposited.
51046_ch13_p141-152.qxd 5/19/11 2:19 PM Page 144
Background
The Burritt Library at Central Connecticut State University is part of a consor-
tial system of five libraries encompassing the four Connecticut State Universities as
well as the State Library. We have approximately 12,000 students and around
another 1,000 faculty members actively using the library. In addition to traditional
holdings the library has extensive Special Collections in Polish as well as digital
materials pertinent to the state of Connecticut and especially the city of New
Britain.
I became the Systems Librarian at Central Connecticut State University in
2005. At the time several competing solutions were used to serve digital objects to
users, and all of the originals were stored in folders on a networked drive. There
was no set schema for finding any of the original objects so that folder names were
the only indicator. There was also no set method of performing checks on data
integrity. Within the Connecticut State University system each library is indepen-
dently responsible for any digital objects outside of the shared integrated library
system and a collection of art images stored in the LUNA system. We also had an
EPrints repository for storing all of our student theses. Part of the Systems Libra-
ian’s responsibility is to administer the servers as well as chart a path forward. To
help accomplish both tasks unifying the disparate sources of data became a top
priority.
As a first step we migrated as many of the digital objects as feasible, the LUNA
144
art images which are shared with another library, into CONTENTdm. CON-
TENTdm is OCLC’s proprietary solution for archiving and display of digital
objects, which in Burritt’s case includes high-resolution scans, video, and text. In
the case of text such as student theses the entire document is uploaded to the
CONTENTdm server and displayed in its native format, such as PDF. The other
types of data stored pose different challenges. High-resolution scans of special col-
lections documents start out as very large TIFF images that are then converted to
JPEG or JPEG2000 images for the web. The question then became what to do
with the very large original TIFFs that were not being loaded onto CON-
TENTdm. A similar situation occurred with our Veteran’s History Project, which
includes interviews averaging two hours with WWII veterans. CONTENTdm
only holds the metadata and points to our streaming server.
Selection Process
As a result of all this valuable digital information leftover, we decided to pursue a
policy of archiving the “digital originals.” This did not necessarily mean an object
had to have been “born digital,” but it was usually a result of hard work on the
part of library staff. As an example, a work might exist in special collections and be
scanned at a high resolution. The original was undoubtedly the work scanned,
but the TIFF also had intrinsic value. The cost of work done had to be factored as
well as the fact that derivative works might be gathered from it, and, in the case of
51046_ch13_p141-152.qxd 5/24/11 8:21 AM Page 145
the displayed object, this was of a much lower quality. Because these items had
value, several solutions were proposed.
The first solution considered was to do nothing. The university’s IT depart-
ment maintained a generous network share of 100 gigabytes (GB) of storage.
When the possibility of storing up to 4 terabytes (TB) with room to grow came up,
the library was quickly given a lesson in the economics of backup. Because of legal
issues as a state institution, Central Connecticut State University has very extensive
and expensive backup procedures. On average it costs our IT department $10,000
a TB for backup. The network share was infeasible.
The next solution considered was OCLC’s Digital Archive product. This was
a natural fit because there was a tie into CONTENTdm. Digital Archive also had
everything the library wanted, including built-in systems management, physical
security, data security, and disaster recovery. Furthermore, the library would not
have to invest in staff training because the procedure is very simple. Unfortunately,
Digital Archive is also very expensive. The library simply could not afford the
costs when dealing with the volume of data being considered.
The solution the library settled on was entirely home grown, using off the shelf
components as well as a combination of low-cost proprietary and open source
software. This was not without some controversy. I have long been a proponent of
open source solutions and had originally proposed the use of a Linux box with
scripts to do everything. Fortunately, the archivist, who has a good computer back-
ground, made a compelling case to try out a solution that seemed at first glance
completely inappropriate: Windows Home Server. 145
Our Solution
Windows Home Server
Windows Home Server is an operating system introduced by Microsoft in 2007
and, as the name implies, was targeted at home users with several computers.
Because of its ease of use and reliability it has gained a great deal of popularity
and has an active community of developers who create add-ins. The vendor from
which we purchased the server, Hewlett Packard (HP), provides some very pol-
ished plug-ins that assist in backup operations as well as system monitoring.
Because these units were designed for home use they do not need as much atten-
tion as a traditional server.
The choice of Windows Home Server as a solution made sense on several
levels. The first was cost. The HP box and 2 TB disks in a RAID array could be
set up much more cheaply than a comparable Linux box. While something along
the lines of FreeNAS using an old computer might have sounded tempting from an
economic point of view, the Windows Home Server box was actually cheaper in
the long run. Another issue was software. Although there are open source solu-
tions available, none of them was as polished and easy to implement in the library’s
51046_ch13_p141-152.qxd 5/19/11 2:19 PM Page 146
IT environment as the Windows solution. The drive was able to be mapped easily,
and remote desktop allowed for easy access to the server even from a Mac envi-
ronment. A full-fledged Windows Server was never considered because of the
complexity of administration and costs involved.
Amazon S3
Available software was another factor as the decision was made to use Amazon’s
S3 cloud computing environment. This choice was made because of S3’s price
and scalability. Software was also abundant for uploading, scheduling, and syncing
of data. Another factor working in Amazon’s favor was that we had a history of
doing business with the company. This is no small matter for a state institution
that can use only approved vendors. Finally, all of the scripts needed to run other
solutions involving data integrity, XML transformations, and fixity checks were
easily available and inexpensive or free.
Implementation
Once hardware decisions were made, we started figuring out a workflow plan to
actually implement the process. Fortunately, we had access to a small Windows
Home Server to test everything out on. Additionally, the Systems Librarian created
a MySQL database based on the requirements for tracking that were specified in
the workflow plan. In addition to the database an HTML interface was con-
146 structed and several MySQL clients tried out before settling on a solution.
At this point it became necessary to write a formal policy to be approved by the
director on how and what would be stored to the cloud, who would pay for it, and
who would do it. The library formed a working group that considered all of the
impacts and wrote up a policy document that spelled out the details. Of particular
importance was the continued funding of this project because there would be
ongoing expenses involved that were sure to increase with the proportional
increase in data. Additionally, the workflow plan was added to so that everyone
involved in the process would know what the procedures were for depositing an
item. Finally, staffing issues were addressed and a long-term commitment of
resources was secured.
The workflow, shown in Figure 13.1, was modeled on OCLC’s Digital Archive
product. A focus was placed on:
• Systems management
• Data security and integrity
• Data backups
Process
Systems management consists of policies and workflows that ensure the data being
stored is always findable by staff and in good shape wherever it is. The workflow
51046_ch13_p141-152.qxd 5/19/11 2:19 PM Page 147
147
starts with a user placing an item or items on the server. This is usually a high-
resolution TIFF. If it is only one or two items, then it is simply a matter of drag-
ging the item to the mapped drive and placing it in the correct folder. This is an
inbox folder that serves as a waiting area until the object can be processed by staff. If
the items are large, consisting of more than 2 GB, then a removable hard drive is
used to ease the strain on the network. The hard drive is then physically attached to
the server and the items are transferred. Once the object is deposited on the server
51046_ch13_p141-152.qxd 5/19/11 2:19 PM Page 148
it is processed by staff. First a MySQL database is populated with data that shows
when the item was deposited, who did the depositing, location, and checksum data.
To capture this data the Burritt Library uses the Library of Congress’s BagIt
system. BagIt is “a set of guidelines for creating and moving digital containers
called bags” (http://www.digitalpreservation.gov/videos/BagIT0609.html). We
use the free Library of Congress Transfer Tools available at http://sourceforge.net/
projects/loc-xferutils/ to create “standardized digital containers,” which include
some sparse metadata and built-in inventory checking. This system works very
well because anything digital can be placed in a bag and a manifest can be gener-
ated to ensure data integrity.
In addition to the checks run with BagIt the library uses JHOVE, the
JSTOR/Harvard Object Validation Environment (http://hul.harvard.edu/jhove/),
to ensure object level validation. While the BagIt format captures data at the level
of bags the individual items in that bag are scanned with JHOVE, which provides
information about the format of an object and its “well-formedness” and validity
that ensures the data has not been corrupted. By employing a series of batch files,
this process is more or less automated. JHOVE file format analysis is quickly
and easily done at the command line. This becomes part of the Preservation Meta-
data or PREMIS. PREMIS is a data dictionary developed by the Library of Con-
gress and “includes semantic units for Objects, Events, Agents, and Rights”
(http://www.loc.gov/standards/premis/v2/premis-dd-2-0.pdf ). This allows us to
define the actual object in a consistent way and not just format details.
148 Several tools exist to facilitate the generation of PREMIS data of which the
library has chosen the excellent Archivists’ Toolkit (http://www.archivist-
stoolkit.org/). The output is a series of XML metadata documents included with
the original objects that look something like this output file from JHOVE:
<?xml version="1.0" encoding="UTF-8"?>
<jhove xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xmlns="http://hul.harvard.edu/ois/xml/ns/jhove"
xsi:schemaLocation="http://hul.harvard.edu/ois/xml/ns/jhove
http://hul.harvard.edu/ois/xml/xsd/jhove/1.6/jhove.xsd"
name="Jhove"
release="1.5" date="2009-12-19">
<date>2010-02-24T16:13:13-05:00</date>
<repInfo uri="C:\test\acpcc_35\bag-info.txt">
<reportingModule release="1.3" date="2006-09-05">ASCII-
hul</reportingModule>
<lastModified>2010-02-24T16:09:29-05:00</lastModified>
<size>70</size>
<format>ASCII</format>
<status>Well-Formed and valid</status>
<mimeType>text/plain; charset=US-ASCII</mimeType>
<properties>
<property>
<name>ASCIIMetadata</name>
51046_ch13_p141-152.qxd 5/19/11 2:19 PM Page 149
A great deal of data can be gleaned from this file, including checksums for
checking data integrity, format information, and dates modified. All of this helps
ensure data integrity when depositing digital objects to the cloud. There are con-
stant updates to these features as cloud-based storage vendors realize the need for
greater precision in long-term data storage to protect against data degradation and
errors in transmission.
After all of the metadata is applied and the object has been tracked it is placed
on the Windows Home Server in a RAID array and then copied to the S3 cloud as 149
part of a regularly scheduled job at 3:00 in the morning when the network has the
least traffic. This schedule was worked out with our IT department for maximum
efficiency. Strong communication with IT is vital. Although the storage is out-
sourced IT is still the network provider, and every effort should be made to make
sure library operations do not affect the overall network adversely.
Ingest $2,400 $0
Internal staff None extra $800 Use current staff, four days/year: $800
requirements ($200/day)
150 able to duplicate most of the features at a fraction of the cost. The estimated cost
for 5 TB of storage with OCLC was $32,250 at the time of this writing. With all
library expenses included, such as staff time and hardware, the total cost for S3
was calculated at $9,362. The majority of that is the recurring annual cost for 5 TB
of $8,610. If your library needs less space the cost would also be less. Table 13.1
shows a breakdown of the costs. The deciding factor for our library was the yearly
storage fee of $26,850.
puting solution from Amazon has worked marvelously and will likely continue to
be used in the future. This is also a highly transferable solution, as the costs in
terms of physical and human resources are remarkably low. As libraries struggle
with budget shortfalls this kind of approach works well in an environment where
the option to “buy it” is too expensive and the option to “build it” is limited by a
lack of technical staff. With just one or two individuals and a modest commitment
of resources, any library can save its precious objects in the cloud.
References
Caplan, Priscilla. 2010. “The Florida Digital Archive and DAITSS: A Model for Digital
Preservation.” Library Hi Tech 28, no. 2: 224–234.
Dougherty, William C. 2009. “MANAGING TECHNOLOGY Preservation of Digital
Assets: One Approach.” Journal of Academic Librarianship 35, no. 6: 599–602.
Library, Information Science & Technology Abstracts, EBSCOhost.
151
51046_ch13_p141-152.qxd 5/19/11 2:19 PM Page 152
51046_ch14_p153-158.qxd 5/19/11 2:19 PM Page 153
Chapter 14
Introduction
The University of Michigan Library moved its research guides content into
Springshare’s LibGuides product in 2009. At the same time that the library was
taking advantage of the benefits of the LibGuides software for authoring and man-
aging the content of the library’s research guides, there was a concern over losing 153
the ability to access, index, and manipulate the information contained in the
guides. Through Springshare’s XML output process guide content was included
in the library website’s search and browse functions and made available to other
parts of the site. Even though the library’s research guides are hosted remotely,
users have access to the content (through keyword searches, tags, and subject asso-
ciations) in other parts of our library website. In this chapter, we discuss our deci-
sion to move to the cloud, what we gained and lost, and how we used existing
tools to maintain a high degree of findability in our local website for the cloud
content.
Background
The University of Michigan Library is one of the largest university library systems
in the United States. The Ann Arbor campus library system holds more than 8.5
million volumes, employs over 500 talented staff, and serves 42,000 students in 19
schools and colleges.
From the mid-1990s through the late 2000s, websites for the University of
Michigan Library’s constituent libraries, services, and departments were locally
managed. Each unit’s site evolved independently from the main library site. To
resolve the resulting confusion of interfaces, library administration launched a
51046_ch14_p153-158.qxd 5/19/11 2:19 PM Page 154
website redesign project in 2008, the goal of which was to remove location-based
content or organizational silos and to present library information as well as web
interactions into a single unified library identity while providing commonality to
the website user experience. During this website redesign project, the librarians
identified a unique category of content that would benefit from a single, user-
friendly presentation: research guides.
Two questions about this content immediately surfaced:
1. Given the large, distributed staff, what tool or set of tools could we use
to manage our subject and research guides (which included research
guides to academic disciplines, guides developed to assist students
enrolled in specific courses, and tutorials on specific software tools and
library databases)? We use the terms “subject,” “course,” and “technol-
ogy” guides to describe these three content types in this case study.
2. If hosted on an external vendor server, how could we connect the valu-
able guide content created by experts to our now centralized library
website for users to discover?
This case study explores the decision process that led to our cloud-based
solution.
Selection Process
Prior to the library website redesign project, we presented guide-like information
154 across many library webpages in many different ways, including a custom-built
PHP application. ResearchGuide, a tool developed at the University of Michigan
Library in the mid-1990s and released as open source software (http://
researchguide.sourceforge.net/) for later adoption by other libraries, accounted
for some of the library’s research guide content. Many library websites reflected
the use of stand-alone HTML pages created by librarians; however, some guide
content was stored on personal web space and not on the library web server.
Our main objective in this project was to select a tool that would simplify the
authoring and user interfaces while at the same time preserving findability to the
guide content throughout the library website. Any guide tool selected would need
to integrate into our Solr-powered search function as well as into our academi-
cally focused and locally developed browse taxonomy, for the purpose of allowing
this unique content discoverable through the website search and browse func-
tions.
As we explored the landscape of options available, we quickly discounted the
homegrown PHP application because it was not sufficiently powerful or flexible to
meet the evolving needs of our content and could not handle Web 2.0-like fea-
tures, such as integrating RSS feeds or embedding video clips. Its administrative
and authoring interface was cumbersome for library staff to use, and the user dis-
play was simplistic and inflexible. We investigated existing modules in Drupal but
51046_ch14_p153-158.qxd 5/19/11 2:19 PM Page 155
Keeping Your Data on the Ground When Putting Your (Lib)Guides in the Cloud
did not find one that provided the necessary flexibility. Wishing to avoid a custom-
built Drupal module, or spending the time and resources revamping existing
applications, we looked for third-party solutions.
In 2008, there was only one viable commercial product on the market. Spring-
share’s LibGuides product (http://www.springshare.com/libguides) offered a
simple authoring environment, enabled sharing of individual content elements
among guides within the LibGuides system, allowed some customization of the
interface, and had an XML output format. This XML output feature was initially
offered as the safety net for libraries, that is, a way for a library to get its data back at
some future date if the library migrated away from Springshare’s product or
Springshare stopped offering it. We saw this as an opportunity to take advantage of
the cloud to manage the guide content while still having access to our data for local
integration and manipulation. In response to a request to enable frequent XML
output (a feature that was not in the company’s short-term plan), Springshare
modified and developed a mechanism to allow for XML output of a library’s
guides on a weekly basis. When needed, the output can be manually refreshed on
demand through a web interface. Here, as in other areas, Springshare’s technical
support group has proven responsive to our needs, confirming the experience
other Springshare customers shared with us when we were exploring commercial
options for our research guide content.
Implementation
At the time of our site redesign, we also dramatically expanded our locally devel- 155
oped subject taxonomy, used in many library-related tools. This expansion meant
that few guides’ content was directly relevant; most of the guides needed revision,
and many subject guides had to be created to fit the new subject taxonomy. Thus,
there was no automatic migration path from the existing tools and webpages into
LibGuides; content migration was almost entirely manual and necessitated library
staff moving or re-creating content in the new tool.
As we developed our Drupal website, we identified areas in which we wanted
research guide content to appear. In particular, we wanted guide content to be fea-
tured in search and browse results and on guide authors’ profile pages. Through
the XML output described earlier, this was now easily achievable.
The metadata associated with Springshare’s LibGuides is sparse; there is
author information, a list of tags, and then the full text of the research guide. The
LibGuides XML output itself is generally unremarkable: The XML is structured
as a representation of the underlying relational database tables. Every week,
Springshare generates the XML as a compressed text file that we download locally
and ingest into our Solr index.
The interesting part is what we do with the XML output. In addition to normal
tagging use, some special tags carry metadata managed within LibGuides but not
expressed in the native LibGuides interface. For example, extra topic categories
51046_ch14_p153-158.qxd 5/19/11 2:19 PM Page 156
associated with a guide (only one is displayed in LibGuides),and our guide classi-
fication (research guide, technology guide, or course guide) are available through
the XML file but not through the web interface. We index the guide content along
with information regarding special tags in our website’s Solr index (see Figure
14.1 for a schematic of the data flows between LibGuides and our website).
Once the data is in our Solr index we can query and manipulate it to integrate
within our Drupal website. The first and most prominent place we integrated con-
tent stored in LibGuides is our search interface. When users search the library
website through the search box in the site’s standard navigation, the results are
divided into categories based on what kind of resource was returned. Thus data-
bases are shown separately from electronic journals or guides, and all of our
resources can be searched from a single starting point.
The second way we integrate guide content from LibGuides is through the
library website’s browse function. Using our locally developed subject taxon-
omy, the browse function allows a user to select a topic from a list to view all of
our resources organized by broad category. Our guide content is findable
through the website browse function thanks to the special tags we included on
the guides.
Keeping Your Data on the Ground When Putting Your (Lib)Guides in the Cloud
The final way we integrate guide content into the library website is in the sub-
ject specialist profiles. Each profile displays a dynamically generated list of each
guide authored by that subject specialist.
Although we feel we made significant gains in the user interface by moving our
guide content to the cloud, there were some minor losses along the way. As with
any remotely hosted tool, the data owners lose some control over user interface
and functionality of the application itself. This is not a criticism of this specific
tool or vendor; it is part of the basic tradeoff of moving content and functionality
off-site.
Of course, there are problems with the data ingestion from time to time. As
mentioned previously, the XML output is a weekly process, which means that
when a librarian edits, adds, or deletes a guide, the changed content is not reflected
in the library website’s search and browse results until the following week. We can
manually download and process a fresh XML file whenever we wish, but the auto-
mated process is available to us, according to our current agreement, only on a
weekly basis.
with phrases like “This guide lists the most important starting points for research
in the area of Economics.” This type of phrasing left the searcher with little context
in which to understand the guide’s presumed content.
Similarly, we also incorporated controlled vocabulary tags into the LibGuides
tags, asking guide authors to use one of several set phrases to describe the kind of
guide (“Course Guide,” “Subject Guide,” or “Technology Guide”) so that we
could label the results accordingly. By setting up rules and standards for guide
authors, we have been able to enhance the findability of LibGuides content in our
site through the search and browse results.
Conclusion
In terms of overall satisfaction with our hybrid of remote hosting of content and
local discovery, we are very pleased with the result. We met our goal to integrate
research guide content into our Drupal website without having to make a sub-
stantial investment of time or resources in developing or customizing a tool of our
own. The user gets the advantage of LibGuides’ ease of use and our website’s
focus on discoverability.
There are benefits of scale, as well. More customers using the same tool means
that when a few customers request a feature that is subsequently developed, it
immediately benefits all. A case in point is the ingestion of LibGuides content in
Serials Solutions’ Summon product announced in late October 2010. While our
library had not requested this, we will be able to benefit from having our
158 LibGuides content available through our Summon implementation. Going for-
ward, we plan to expand the integration of the guides content into course-, depart-
ment-, and discipline-specific pages, as well as to integrate LibGuides content into
the University of Michigan’s Sakai-powered course management system.
51046_ch15_p159-164.qxd 5/19/11 2:23 PM Page 159
Chapter 15
Introduction
This chapter highlights one file-sharing program in particular, Dropbox. The pro-
gram has generated a great deal of buzz by users, and, while it certainly seems to be
well liked by many single-source users, the question remains as to whether or not it
is well poised to be used within the context of libraries. The following case study
will show how the librarians and staff at Murray State University library used the
service and will seek to offer input for libraries considering adopting the service. 159
This chapter will give an overview of the service and how the embedded librarians
within our library used it, as well as highlight other options we might have taken
on in the cloud computing world.
Background
Murray State University is a midsize university with over 10,000 FTE students in
Western Kentucky. The University Libraries see on average 3,500 students pass
through the doors of the main library and use services every day. Because the
library serves over a third of the entire student population daily, there is a tremen-
dous strain on resources. We have two computer labs consisting of 74 computers
total and a collection size of over 700,000 print materials. While it could easily be
said that many of the students who use the library each day are using it as a central
location to study, check e-mail, and grab a quick soda, this influx of students helps
to determine how library materials are managed and allotted. The library Refer-
ence Department has watched as their traditional roles have changed, with focus
shifting from providing in-depth research on a wide variety of topics to simply
acting as the focal point of the library where they often become the go-to depart-
ment for everything from research to help when a printer is faulty. Embedded
librarians as part of the Reference Department have struggled to find the best ways
51046_ch15_p159-164.qxd 5/19/11 2:23 PM Page 160
to balance their dual roles as frontline faces of the library and as teachers. Primar-
ily, each department within our library system communicates to the others through
the library listserv or shares files through the R-Drive, our local shared network
drive. Although initially librarians were intrigued by the uses of Dropbox, few saw
the benefits to using Dropbox over the R-Drive.
Selection Process
Overview of Dropbox
As of this writing, Dropbox can still be considered a relatively new program. Cre-
ated in 2007 by Drew Houston, it has quickly grown to make a significant name
for itself in the cloud computing market as a software as a service (SaaS) provider.
SaaS is software that is typically provided on demand, within the cloud, and pay-
ment is typically open source or pay as you go. The way the program works is
simple: users download a program onto any computer or device they want linked
into the Dropbox system. For most people this link will be to their home and
work computers, though others have chosen to link their smartphones to Drop-
box as well. This program creates a Dropbox file folder on the computer that
updates across all linked computers and devices. If a file is added to a Dropbox
folder on a personal computer, the same file will then be added to the Dropbox
folder on that employee’s work computer as well. This creates local copies of
each file in every location that we have specified to be synced with Dropbox. The
160 concept of Dropbox holds up nicely with the LOCKSS (Lots of Copies Keep
Stuff Safe) theory of library science because we end up with as many copies as
computers that are regularly used to link to a particular Dropbox account or
folder.
There are two different types of Dropbox accounts: users can create a free
account giving them access to 2 GB of data, or they can pay for one of two tiered
levels of data storage. For $9.99 a month, users can get up to 50 GB of data stored
on a Dropbox node, and for $19.99 users can receive access to up to 100 GB of
stored data. Although it is not an invite-only service, similar to many Google appli-
cations in their Beta stage, Dropbox does reward its users who bring new users
to the site. Free users of the website have the potential to receive up to 8 GB of
data storage if they refer enough friends to the service.
Dropbox’s web interface is very clean and minimalist. Files and folders are
listed out in easy formats similar to most Microsoft operating systems. There are a
handful of tabs at the top row that help you navigate, but these are few and easy to
understand. Users log in to the site with their e-mail and password, and then the
first thing they are greeted by is their files. There is no need to make several clicks
and wade through pages to get to the proper files.
Changes to the system can easily be tracked through the “Events” tab on the
web interface of the site. Here is where the user can see what has been added and
51046_ch15_p159-164.qxd 5/19/11 2:23 PM Page 161
edited along with the day. There is a structural difference to this within the two
versions of paid and unpaid accounts. Free accounts can only see changes made to
the system within the past 30 days. Anything stored before 30 days cannot be
accessed. However, paid subscribers at any tier level have access to all edits and
changes in files through their system, Pack-Rat. This is viewable only through the
web interface. If users upgrade to a paid account from a free account, they do not
gain automatic access to their previous edits. Rather, users only get edits in the
past 30 days and onward. It is not a retroactive program, and so it takes effect only
from the moment you upgrade to a Pro account rather than from the beginning of
your use of Dropbox. Having access to edits made to documents and the system is
a great advantage, as it allows you to see older versions of documents as well as to
(sometimes) see who made those edits.
Implementation
Embedded librarians seemed perfectly poised to use Dropbox. At Murray State
University there are currently six embedded librarians, each frequently traveling to
51046_ch15_p159-164.qxd 5/19/11 2:23 PM Page 163
satellite campuses, and half of the embedded librarians maintain offices in their
subject-specific department on campus. The embedded librarians travel a great
deal, but they still need to access common files across all accounts, so Dropbox
seemed like a natural fit. Although many of the librarians have the option to carry a
laptop with them to their alternate offices and to the classrooms when they teach,
Dropbox has provided another way to readily access their information when they
are outside the typical library environment. The embedded librarians have used
Dropbox most when they are off campus at conferences or preparing classes at
home. It spares them the extra step of having to log in to the university VPN, as all
they have to do with Dropbox is ensure that files have recently been updated in
their Dropbox folder, and then they can access the same file that is stored on their
work computer. Typically, when embedded librarians are teaching a class outside
of the library, they will simply log on to the web interface of Dropbox and then
download the lessons or presentations that they have already prepared and loaded
onto the Dropbox server earlier. It quickly downloads, and their presentation is
ready to teach in seconds.
At Murray State we opted for the free version of Dropbox, so the amount of
stored data available to librarians and staff was comparable to that of a personal
USB drive. One unexpected debate sprang from whether every librarian should
have his or her own personal Dropbox account or whether there should just be a
general Murray State account set up. If librarians were to create personal accounts
they would have more space set up for just their work, but if a general account was
set up for the entire department then everyone would have equal access to every- 163
one’s shared documents without the trouble of setting up unique sharing settings
for every individual account. Eventually it was decided by the head of Reference
that a general account would be set up with departmental folders. Then individual
librarians created their own accounts that they used to get shared access to those
folders.
notifications when edits or changes had been made to a document. Although those
edits and changes would be noticed through the “Events” tab or by opening up the
document itself, it was not something that was immediately apparent.
Overall the librarians were very pleased with the service and have continued to
use it for several other small projects; however, the only major hurdle we seemed to
encounter with Dropbox was that we already had an internal server that most
librarians were used to using. The service became truly useful only when the
embedded librarians were outside of the library teaching in other classrooms.
Lesson plans and assignments were then easily accessed through the service. It is
definitely a service that libraries should consider adopting into their own pro-
grams, as it adapts easily to the many needs of data storage that a library might
have. The service would be recommended to anyone considering entering the
cloud, as it offers a wide range of possibilities for temporary data storage that seem
very well suited for the small or medium-sized library. As a personal service, it
would be definitely recommended that individual librarians take advantage of this
service if they have a need to access files in multiple locations. This is a highly
accessible service that is ideal for the many different ways that libraries and librar-
ians work.
Conclusion
Cloud computing seems to have many unique advantages over local servers and
offers new methods of access to information for both students and librarians. As
164 many people have already addressed, cloud computing has some very serious legal
and security questions that may be off-putting to libraries before they attempt to
store what can be massive amounts of data online. Although there are public
options to Dropbox, it seemed that primarily when our librarians were using the
service it became almost an internal application that many used to share files from
coworker to coworker(s). Rarely were public documents kept on the service, and
so we had very few privacy concerns to worry about.
If our library were to continue heavily using this service, we would probably
opt for one of the paid versions of Dropbox, simply because 2 GB of space gets
taken up quickly when you have multiple people using the service for varied pro-
jects. Probably the hardest part of using it was just in learning to break out of our
old habits and remembering that we had the option to use it. There was a very
minimal learning curve to using the site, and it continues to be used by many of the
embedded librarians, who see it as an effective tool for organization across loca-
tions.
51046_ch16_p165-172.qxd 5/19/11 2:23 PM Page 165
Chapter 16
Introduction
The instruction librarians of the Ursula C. Schwerin Library at New York City Col-
lege of Technology were faced with an enviable problem: a vigorous and growing
instruction program, a small core of instruction regulars, and one 30-seat e-class-
room, the scheduling for which required great promptness and accuracy. A trans-
parent way of creating a schedule for the library’s electronic classroom, the room in
which most library instruction occurs, was needed. At first librarians moved from a 165
longhand, paper-based, and Excel-dependent schedule to the calendar feature of
the college’s e-mail system, GroupWise. When GroupWise mysteriously and spon-
taneously deleted dozens of scheduled classes from all library faculty members’ cal-
endars, and campus information technology services could neither recover the data
nor help us learn how to prevent this problem from recurring, other online shared
calendar tools were investigated. Some librarians were already using Google Docs
and Gmail, so Google Calendar, a browser-based application available to any
Google account holder or Gmail user, was easily within reach. Google Calendar
proved to be a useful tool for e-classroom scheduling, and now the library’s faculty
and staff depend on ten different Google Calendars to communicate about classes,
meetings, annual leave, and the reference desk schedule.
Background
The Ursula C. Schwerin Library serves the students, faculty, and staff of New York
City College of Technology (City Tech), the largest public baccalaureate college of
technology in the Northeast. Enrollment in fall 2009 surpassed 15,000 full-time
and part-time students, according to the City University of New York’s Office of
Institutional Research and Assessment (CUNY OIRA, 2009). According to the
U.S. News & World Report’s 2011 college rankings, City Tech is the second most
51046_ch16_p165-172.qxd 5/19/11 2:23 PM Page 166
diverse regional college in the North (U.S. News & World Report, 2010). City Tech
is a college in the CUNY system, which consists of 25 colleges throughout the five
boroughs of New York City and enrolls nearly 260,000 degree-seeking undergrad-
uate and graduate students. Since 2008, the library instruction regulars at City Tech
have grown from a group of three instruction and reference librarians to a group of
five librarians. Two new positions to support library instruction and information
literacy were created and filled: a Coordinator of Information Literacy and an
Instructional Design Librarian. As the instruction regulars, we take responsibility
for planning, scheduling, and delivering the bulk of the library instruction program;
our department chair determines how many classes each instruction regular should
be assigned, with consideration given to tenure status and other responsibilities.
At City Tech, every student is required to pass English Composition I, and
each section of that course visits the library for one library research lesson. In addi-
tion to teaching library research lessons for all sections of English Composition
I, librarians coordinate other requests for library instruction, passing them along to
librarian subject specialists as appropriate. All City Tech librarians are responsible
for library information literacy instruction for the departments for which they are
subject specialists. We also guide technology implementation in the electronic
classroom and recommend software and hardware upgrades for purchase when
we are fortunate enough to have a budget to do so.
The library’s instruction program was growing quickly and demands on our
teaching facility—one smart classroom equipped with 30 student computers, two
166 projectors, and an instructor computer—were growing. Not only were there
increasing numbers of English Composition I sections as college enrollment grew,
there was increased interest in library instruction from departments all over
campus, especially Speech, Nursing, Health Services Management, and Hospital-
ity Management. Library faculty also formed a workshop group to plan and deliver
workshops to students and faculty. The Information Literacy Librarian, supported
by the department’s curriculum committee, developed a credit-bearing course that
can be taken to satisfy a core course requirement in the area of communications for
two-year and four-year degree curricula. With multiple departments demanding
library instruction, and the library itself expanding its instructional offerings, there
was a great need for simple, quick, and transparent e-classroom scheduling that
all librarians could easily access and update.
Selection Process
Before we switched to an online shared calendar tool, the instruction regulars (first
three, then four, and now five librarians) had established a procedure for scheduling
library instruction. Each semester, the most senior instruction librarian obtained
the English department’s schedule of classes and used a paper calendar template to
write out a schedule in longhand. The handwritten calendar (Figure 16.1) was sub-
mitted to the department’s Administration Office Assistant, who transcribed the
51046_ch16_p165-172.qxd 5/24/11 8:23 AM Page 167
167
handwritten schedule into Excel. The Excel document was printed and distributed
to all librarians with instruction duties, who would take a few days to check their
teaching schedules against their existing calendars, reference desk schedules, and
other department and college commitments. On approval of library faculty, the doc-
ument was shared with the English department. Each time changes were made to
the library instruction schedule, the Excel document had to be corrected and redis-
tributed. All of the instruction regulars, however, needed to be able to update the e-
classroom schedule and to communicate changes quickly to all librarians.
GroupWise
When librarians began to look to an online solution, they used the calendar event
feature in GroupWise, the college-wide e-mail client. The ability to manage
changes rather than working through one very busy staff member meant that
schedule updates and changes were made quickly and with very little effort. With
help from the college IT support, we created a new user account in GroupWise,
LibraryClassroom, to which the instruction regulars had access. The Posted
Appointments feature was used to create appointments on LibraryClassroom’s
51046_ch16_p165-172.qxd 5/19/11 2:23 PM Page 168
Google Calendar
Although troubled by the mysterious failure of GroupWise, we were reluctant to
return to a print calendar. A few librarians already made use of other Google appli-
cations, including Gmail and Google Docs, for personal and professional com-
munications and productivity. Because use of Google Calendar requires only a
browser and an Internet connection, and does not require investment in software
or storage media, moving the library instruction schedule to Google Calendar was
an easily implemented solution to an urgent problem. The transition from Group-
Wise to Google Calendar went quickly; because we had kept careful records on
paper, and had retained the class schedule from the English department, we were
able to reconstruct all e-classroom events within a few days. We kept a GroupWise
backup for a short time.
The introduction of Google Calendar bolstered a shift in attitudes about
scheduling. As the transition from paper to Excel to web-based calendar tool has
transformed the way we operate daily, how we conceptualize the task of scheduling
168 has changed. Now the Google Calendar is the living document, and the spread-
sheet generated at the end of each semester is the record, whereas before Google
Calendar, the spreadsheet had to be both the living document and the record.
Google applications, including the calendar, are imbued with “cloudiness”; they
are meant to be transparent and accessible tools to facilitate sharing. GroupWise is
more of a personal application with little to give a sense of a shared space for com-
munication or documentation. When switching tools, we underwent a collective
shift in attitude about ownership, privacy, and empowerment.
Implementation
When we first began to use Google Calendar, one librarian created a calendar using
her own personal Google account and shared it with the other instruction regulars.
It soon became clear that we would benefit from being able to view the reference
desk schedule, colleagues’ on- and off-campus meetings and appointments, and the
annual leave calendar simultaneously. At that time, library administration created a
Google account called “citytechlibrary.” Only the department chair, the adminis-
trative services librarian, and the administrative assistant have access to edit calen-
dars created by this account: Library Faculty Annual Leave, College Cultural
Events, College Office Assistant Absences, and Public calendars (the Public calen-
dar was developed to embed into our website or blog and promote library events).
51046_ch16_p165-172.qxd 5/24/11 8:23 AM Page 169
All library faculty and the department’s systems administrator have rights to create
and edit all other calendars: College Meetings, Department Meetings, Professional
Meetings, Reference Desk, Small Library Classroom, and E-Classroom.
The learning curve was steep at first. Librarians and library staff who did not
already have Google accounts had to create them; there is no way to view a Google
Calendar that is not made public without logging into Google accounts. Instruc-
tion regulars offered one-on-one training to all library staff and faculty on the cre-
ation and maintenance of calendar events. A protocol for entering data for library
classes was quickly established; a consistent pattern greatly enhanced the search
functionality. Now any librarian can view instruction responsibilities, reference
desk shifts, or scheduled meetings from anywhere the Internet can be accessed.
The Information Literacy Coordinator early on recognized the importance of
data backups and makes it a practice to regularly export the Google Calendar into
a zipped file and save it on the college’s Internet-based file storage system. Data is
stored in iCal format. So far, we have not had to restore data from a backup zipped
file; data seems secure in Google Calendar. As shown in Figure 16.2, Google Cal-
endar allows multiple calendars to be viewed at the same time.
Google Calendar offers some functionality that the college’s version of Group-
Wise does not: the ability to set e-mail or SMS reminders for calendar events for
early morning classes, evening or weekend classes, or classes needing special
preparation or setup. GroupWise offers pop-up reminders, but only in the desktop
client, not in the web version.
Choices made about Google Calendar options included managing sharing
and write rights. When the department adopted the instruction regulars’
calendar techniques, editing rights were extended to all librarians for most
calendars.
Conclusion
Migrating to shared online calendars for library instruction scheduling and for
most other aspects of library work life has changed how some library faculty use
Google Apps for productivity, including shared documents for collaborative writ-
ing projects and forms to survey classroom faculty about library instruction. A
number of librarians now make use of Google Calendar and other applications for
personal and other professional uses. At some point we may require more func-
tionality and interactivity than Google Calendar can offer. We have learned that,
when necessary, we can engineer what some might have regarded as a temporary
fix into a long-term, hard-working solution.
References
CUNY Office of Institutional Research and Assessment. 2009. “Total Enrollment by
Undergraduate and Graduate Level, Full-Time/Part-Time Attendance, and College.”
Updated March 26, 2010. http://owl.cuny.edu:7778/ENRL_0001_UGGR_
FTPT.rpt.pdf.
Google. 2010. “Google Privacy Policy.” Last modified October 3. http://www.google
.com/privacypolicy.html.
U.S. News & World Report. 2010. “Best Colleges: Racial Diversity: Regional Colleges
(North).” Best Colleges 2011. Accessed October 12. http://colleges.usnews
.rankingsandreviews.com/best-colleges/bacc-north-campus-ethnic-diversity.
172
51046_ch17_p173-180.qxd 5/20/11 8:35 AM Page 173
Chapter 17
Introduction
Google Forms offer librarians an array of opportunities to collect information from
reference patrons and from students engaged in information literacy instruction.
Part of the Google Docs suite of software as a service (SaaS) products, Google
Forms have been recognized by librarians as an inexpensive and easy-to-use survey
tool (Travis, 2010). Beyond surveys, libraries can use Google Forms to receive ref-
erence questions, and Google Forms offer a paperless alternative to the in-class
worksheet often used in information literacy exercises. This chapter describes how
Google Forms have been used in an academic library’s reference and instruction 173
services.
Background
The University of Wisconsin–Eau Claire is a public institution primarily serving
undergraduates, with enrollment of more than 11,000 students. The McIntyre
Library at UW–Eau Claire is a heavily used virtual and physical space. The library
employs a variety of web-based instructional tools, including LibGuides, for gen-
eral and course-integrated research assistance. Virtual reference—chat/IM, text
message, and e-mail—is part of the McIntyre Library’s standard reference service.
“Chat with McIntyre Library” links are displayed in electronic resources, and
chat/IM widgets are embedded in LibGuides. Virtual reference and web-based
research guides are hardly unique among academic libraries; however, these tools
are situated in the conceptual framework described by Lorcan Dempsey (2008:
111) as being “in the flow.” As a library, we are working to adopt tools that enable
us to connect with patrons within their research environment and at the point in
their workflow where they need our assistance. Furthermore, because our reliance
on web-based instructional tools and services is growing, we are actively searching
for tools that will help us create “safety nets” at points where patrons become con-
fused or frustrated by library resources (Veldof, 2008).
51046_ch17_p173-180.qxd 5/19/11 2:23 PM Page 174
Other survey tools had the potential to meet the needs I had identified, but
the free or “basic” products often lacked a single critical component. Marie and
Weston (2009) succinctly review several online survey tools, including Survey-
Gizmo, Survey Monkey, and Zoomerang. I considered the no-cost version of each
of these tools, and I was especially impressed with SurveyGizmo, which offers
Java script and HTML iframe tags for embedding the form. In a literature search
about using webforms in higher education and libraries, Travis’ (2010) description
of a medical library’s use of Google Forms for surveys convinced me that Google
Forms is a sufficiently flexible product. Meanwhile, Gerhinger (2010) described
integrating Google Forms into active learning exercises in computer science
instruction at North Carolina State University. Based on these comprehensive
assessments and my own experimentation, I concluded that Google Forms com-
bined the qualities I sought, including a spreadsheet, into a single program that I
could access anywhere.
Google Form (or to begin any other Google Doc). To implement Google Forms, I
did not need to establish and keep track of another web-based product account,
because my library already has a Google account for Google Voice and other ser-
vices. Although a Google account is required to set up a Google Form and track
the data collected, those who fill in the form do not have to log in or maintain a
Google account.
To create a Google Form, a user must simply log into Google Docs
(http://docs.google.com) with a Google account, select “Create New,” and select
“Form.” A new Google Form is a blank canvas and each field of a new form is
clearly labeled to indicate its purpose. Clicking on a specific field allows the form
owner to edit that field. New questions can be added to each form by clicking
“Add item.” Selecting from the “Question Type” dropdown menu automatically
alters the format of the question, depending on the question type selected.
A completed Google Form can be distributed in three ways. The form owner
can e-mail it with the “Email this form” button that appears when the form is in
edit mode. A unique URL appears at the bottom of the form when in edit mode
next to the phrase, “You can view the published form here”; this URL can be
copied and pasted for distribution via e-mail or as a link on a website. Finally, by
clicking on “More Actions” within edit mode, the form owner can access an
HTML iframe tag, which can be pasted into the source code of a website.
Implementation
176 When I set up Google Forms, I first built a form with a set of prototype questions.
Colleagues and student employees at our reference desk “road tested” the forms,
offering immediate usability insights and some suggestions for revision of format-
ting, question wording, order, and type. Following these revisions, I copied the
HTML iframe tag from my form and pasted it into a LibGuides content box.
I first tested Google Forms in two upper-division history seminars in which
students were beginning a significant year-long research project. In collaboration
with McIntyre Library’s Special Collections and Archives, I developed a LibGuide
for the course and embedded a reference question webform into the LibGuide
(Figure 17.1). I met with both seminars for 50 minutes each, highlighting the
course LibGuide, and encouraging students to submit reference questions using
the form embedded under the “Ask a Question” tab of the LibGuide.
After teaching both history seminars, I eagerly awaited receipt of a question
submitted via Google Form. I did not expect that every student from the history
seminar would choose to ask a question in the Google Form but hoped that the
format was a desirable option for some patrons. In the two weeks that followed
my presentation to both seminars, I received nine reference questions from semi-
nar students. Four questions were submitted via Google Forms, three students
asked questions in person in my office, one question was submitted via e-mail,
and one patron used chat/IM to request assistance. The nature of the questions
51046_ch17_p173-180.qxd 5/19/11 2:23 PM Page 177
177
submitted via Google Forms was similar to reference questions received in other
modes. For example, one student asked how to locate historic census data, while
another student sought help identifying appropriate search terms related to the
eighteenth-century fur trade. Since piloting the form in these history seminars, I
have embedded similar forms in research guides designed for other courses, occa-
sionally changing the wording and format of the form fields to adjust to the par-
ticular research needs of the students in the course.
Data submitted through a Google Form is collected in a single Google Spread-
sheet linked to the Google Form and accessible via Google Docs. Seamless data
collection enables instant comparison of questions received from students who
51046_ch17_p173-180.qxd 5/19/11 2:23 PM Page 178
have attended a library instruction session or used a research guide. This data
allows me to look for patterns that may indicate that the students are having diffi-
culty with particular research concepts or library tools.
Google Forms also have applications for interactive classroom exercises. In
information literacy instruction designed for an English composition class, I cre-
ated a classroom activity to be completed in a Google Form embedded in the
course LibGuide. During a 50-minute session, students are introduced to strate-
gies for finding books and articles, and they compare their experiences with dif-
ferent library interfaces. When students apply their learning during each class
activity, they fill out a brief form describing what they find.
After students submit the form I view the data they submit in a Google Spread-
sheet. Although I believe assessment of student learning should extend beyond
students’ ability to find a book or an article, the data collected in these web-based
worksheets enables quick assessment of whether students succeeded in executing
the in-class activity. Asking students to fill out a web-based form is easy to inte-
grate into the “flow” of a classroom activity that primarily utilizes computers and
the Internet. Crucially, each student must have access to a computer in the class-
room in order to access and fill out the form.
Based on successful implementation of Google Forms, future uses of webforms
for reference questions will include linking the forms to course pages within D2L,
the course management system used at UW–Eau Claire. Gerhinger (2010) pro-
vides a thorough overview of how Google Forms can be used in classroom instruc-
178 tion, both for engaging students in active learning exercises and for collecting
opinions or answers to factual questions. McIntyre Library does not currently have
access to clickers for instruction; however, one might imagine a Google Form that
a librarian could use to conduct polls and collect other data from students who
might otherwise use clickers. Clearly, Google Forms present many possibilities in
instruction.
Students using the forms to submit classroom exercises have not revealed
any frustration with the tool. However, one obvious barrier to using Google
Forms as an instructional tool is that a webform may not fit into the “flow” of a
classroom activity. For example, engaging students in small-group discussions
about search results may not call for completion of a webform. Similarly,
Google Draw (or another drawing product) may be a more appropriate tool for
creating a Venn diagram to compare and contrast reference sources. However,
Google Forms are a highly functional tool for eliminating paper-based work-
sheets in classroom settings, and McIntyre Library patrons have demonstrated
that Google Forms are useful in the menu of options for asking a reference
question.
Conclusion
When I began the hunt for a versatile webform, I wanted a product that was free
and easy to build, customize, embed, and distribute. I was looking for a product
with a variety of question types and with seamless data collection options. I looked
to many online survey tools, but Google’s SaaS product suite met each of these
needs with Google Forms. The customization features of Google Forms have
enabled me to create forms that do not inhibit patron questions with the character
or word limits common to many online surveys. This has been particularly advan-
tageous in instructional settings, because patrons are able to fill out a form just as
they would an in-class worksheet. Implementing Google Forms in the McIntyre
Library has shown that libraries may have uses for webforms beyond collecting 179
survey data. Google Forms’ ease of creation and customization, combined with
low cost and no software installation or maintenance requirements, should be
attractive features to librarians looking for paperless classroom assessment options
and for new ways to connect with users.
References
Dempsey, Lorcan. 2008. “Reconfiguring the Library Systems Environment.” portal:
Libraries and the Academy 8, no. 2: 111–120.
Gerhinger, Edward. 2010. “Teaching Interactively with Google Docs.” Paper presented
at the Annual Conference and Exposition of the American Society for Engineering
Education, Louisville, KY, June.
Google. 2010. “Getting to Know Google Docs: System Requirements.” Google. Accessed
November 1. https://docs.google.com/support.
Marie, Kirsten L., and Janine Weston. 2009. “Survey Says: Online Survey Tools for
Library Assessment.” Library Media Connection 28, no. 2: 50–52.
Travis, Lisa. 2010. “One of Many Free Survey Tools: Google Docs.” Journal of Electronic
Resources in Medical Libraries 7, no. 2: 105–114.
Veldof, Jerilyn R. 2008. “From Desk to Web: Creating Safety Nets in the Online Library.”
In The Desk and Beyond, edited by Sarah K. Steiner and M. Leslie Madden, 120–134.
Chicago: Association of College and Research Libraries.
51046_ch17_p173-180.qxd 5/19/11 2:23 PM Page 180
51046_ch18_p181-186.qxd 5/19/11 2:23 PM Page 181
Chapter 18
Introduction
Custom social networking sites provide an environment that supports engagement
with new ideas and resources. Cloud-based software as a service (SaaS) offerings
like Ning and Wetpaint provide easily customized and implemented virtual space
that fosters community and participant conversation. These services can also sup-
port collaborative, interactive teaching and learning. This chapter details how such
SaaS social networking sites were used to meet the need for asynchronous training 181
for staff in a theological library, resulting in the creation of a model that was later
adapted for use by teaching faculty.
Background
Union Presbyterian Seminary is a small, stand-alone academic institution offering
five masters degrees and two doctorates. Within the library there is one systems
librarian, covering all internal information technology needs, and a campus Tech-
nology Services department, composed of four FTE responsible for all other com-
munications infrastructure, support, and services. During the timing of both
projects discussed, the campus was experiencing financial constraints. There was a
need, from the projects’ outset, not only to outsource the hardware and systems
administration functions made available by “renting” virtual server space, but also
to have these automatically bundled into a provided, supported, stable web-enabled
environment. Planned goals for this environment included alleviating the need for
local system administration, authentication, programming, storage, or server-based
software installs. The environment also needed to provide remote file uploads,
hosted web tools, a browser-based interface, stable service “up time,” and support.
The financial and staffing realities of the institution lead project leaders to pursue a
cost-effective, cloud-based solution with these goals for both initiatives.
51046_ch18_p181-186.qxd 5/19/11 2:23 PM Page 182
Training Series
In the spring of 2010 plans moved forward to offer a training series for staff in the
library. The impact of requiring library staff to move among many individual web
tools was discussed. The decision was made to search for a single cloud-based
platform that would combine as much of the functionality desired for the project as
possible in one place. The hope was that this decision would lower staff anxiety
and resistance concerning technology training and encourage all communal com-
ments or conversation to occur in a single location. A static Welcome page and
the capability for all members to be able to give their responses to the instructional
modules as blog posts were preferred. The training series also called for partici-
pants to be able to experience instant messaging, discussion forums and wikis,
along with other Web 2.0 tools.
Summer Course
Simultaneously a member of the faculty was searching for a way to improve the
at-home sessions of her course. The course, focused on teaching with technology,
was part of the campus’ Extended Campus Program (ECP). ECP provides a means
for working adults to pursue part-time graduate study leading to a masters degree
in Christian Education. Students engage in study at home and attend an intensive
session on campus. Course materials are usually delivered through use of the insti-
tutional course management system. This system supports transmission of mate-
rials but does not provide a platform for an interactive, dynamic learning
182 environment for the at-home portion of the course. The experience of the library’s
training series project leaders with cloud-based custom social networking plat-
forms for asynchronous learning offered possibilities for a more collaborative
model of teaching and learning in the ECP while also modeling the use of such
tools for students’ future professional work.
Selection Process
As the search began for a single, feature-rich, cloud-based platform to fit the insti-
tution’s goals, as mentioned earlier, a list of desired features was established. Along
with basic ease of use for participants, key capabilities were a static Menu or Wel-
come page, built-in instant messaging, discussion forums, timed release of con-
tent, individual participant blogs, built-in wiki, standard GUI editor, the ability to
easily embed content/widgets from other sites, nominal cost, and privacy control.
After some initial searching and conversations with associates at other institutions,
the options were quickly narrowed to two custom social networking sites: Ning
and Wetpaint. Each cloud-based platform offered unique advantages and disad-
vantages in the ways that they were structured.
The custom social networking site offered by Ning, for example, provides two
places where almost all content is collected for sharing: the opening Welcome page
and the individual member page, similar to other social networks’ individual,
51046_ch18_p181-186.qxd 5/20/11 8:36 AM Page 183
member profile page. Outside of these two locations within the site all content is
pooled by format type: blog post, discussion forum entry, uploaded picture, or
uploaded video. The individual piece of content is clearly associated with the user
profile that has created or uploaded it. The ability for participants to comment is
available for all format types. This universal ability to comment on material was
quite important for each project.
The custom social networking site provided by Wetpaint offers most of the
same sought after functionality but delivers it differently. Wetpaint also offers a
default Home menu. It shows subpages, new activity, and members, which are dis-
played as widgets along the page sides. Wetpaint lacked, most importantly, inte-
grated blogging. Instead of blogs as an available format type, Wetpaint uses
“pages,” operating like moderated wikis that have multiple assigned writers. This
structural design led to a very clunky interface and caused concerns about partic-
ipants’ potential confusion regarding how to author new content.
One of the goals for a cloud-based environment for both projects was either a
free or low-cost set of services. When both custom social networking sites were
initially reviewed as solutions, both were free but displayed advertisements in
order to generate revenue. Wetpaint’s design for displaying these advertisements
was much more aggressive. The distinction in how each custom social network
site managed new user content, along with Wetpaint’s very heavy use of advertise-
ments, drove the decision to select Ning.
It is worth noting here, however, that midway through the initial project,
during the training series for library staff, Ning changed to a completely fee-based 183
model. Shortly after, Wetpaint chose to focus their development on fan sites for
TV or movies target markets. More will be said later about the impact of these
changes, but they only reinforced the original choice. Ning’s price for an annual
subscription, as of May 2011, was $19.95 for Mini, $199.95 for Plus, and
$499.95 for Pro. Wetpaint’s Premium Services offering was $239.40 in Novem-
ber 2010.
Implementation
Training Series
In May 2010 the Web 2.0 introductory series for library staff was created using
the selected Ning social networking platform. The custom social network that was
designed used all the following components: blogs, forums, photos, and chat. One
of the key features, wikis, was not available from Ning. Other resources outside of
the custom social network had to be relied on for those needs.
Each learning unit was developed as a blog post by one of the project leaders
and linked to the splash page’s menu as an embedded hyperlink in the unit’s title.
These individual learning units were timed to release to participants on set dates,
so the series could follow a planned schedule that also included “open lab” times
51046_ch18_p181-186.qxd 5/19/11 2:23 PM Page 184
in the library’s media lab when project leaders would be available to answer ques-
tions and assist staff experiencing issues. Participants were then issued invitations
to join the custom social network that was called “7 Signposts to Library 2.0.”
Their activity as they progressed through the series generated individual blog
posts, threaded discussions in the forums, and exposure to instant messaging
using the network’s built-in chat feature. Other activity occurred outside of the
site but was then discussed in reflection pieces created as blog posts.
Summer Course
The course’s custom social network utilized the library training series’ Ning site as
a template. Registered students were invited to participate and were introduced
to the work of the course through the main page of the social network. This page
included information regarding course purposes, methods, and assignments, with
a link to the syllabus. A sidebar depicted members and detailed course activity.
Content material presented perspectives on the weekly topic, with minimal
reliance on a printed course textbook. Exploration of online sources led to arti-
cles, websites and video/audio resources on educational practice, Internet usage,
and teaching in religious education, which became embedded links in the course
site. The Ning blog, associated with the support subsite, directed the project lead-
ers to valuable information and examples of networks that specifically use
Ning in education (http://education.ning.com; Classroom 2.0, http://www
.classroom20.com/).
184 After engaging the materials, students were invited to experiment with web
tools such as Flickr (http://www.flickr.com/) and VoiceThread (http://
voicethread.com/). Assignments asked students to discuss the content and their
experiences with technological tools through blog and forum posts, components
of the Ning system.
sion was made to stay with the network that had already been created. It was felt
that too much effort had been invested in the series to transition. This also avoided
forcing participants to move to another learning platform.
The most significant disappointment was communication based. Participants
expected to receive e-mail notifications when others added content to the train-
ing series’ Ning network. In initial tests among the project leaders, this had been
successful. Once the project launched, the participants did not receive such e-
mail. It was initially thought that the issue may have arisen because network mem-
bers had not been required to “friend” their colleagues. Testing proved this
assumption to be false. The root of this particular issue was never found. Attempts
to contact Ning’s support staff, now that the institution was a paying customer,
failed. This occurred during the period that only support tickets associated with
billing and subscriptions were receiving prompt attention. This unexpected issue
left project leaders with no choice but to rely on using “all member broadcast”
messages as network administrators. This did not make up for individual partici-
pant notification when new content and comments had been added by other par-
ticipants, a reminder that had been counted on to help drive conversation on the
networks. Ning, therefore, met 11 of 12 of our goals for a cloud-based environ-
ment and provided nine of ten key features.
Summer Course
The course use of Ning as an educational platform was successful. Students gained
both an academic knowledge of teaching with technology and technical knowl- 185
edge of Web 2.0 tools. Ning had partnered with Pearson to offer a free “Ning
Mini” specifically developed for educational purposes, which was requested for
the course’s social network. Despite immediate registration, notification was not
received within the term.
The most important assessment focuses on the collaborative learning environ-
ment of the custom social network. The process of learning moved from transmis-
sion of information to collaborative engagement. The changing roles of the
instructors serve as a metaphor for this educational shift. Initial consultation
among project leaders involved a conference between two experts. The librarian
colleague contributed knowledge and experience of social networking as an edu-
cational platform. He also contributed the necessary technical skills that supported
the use of Ning. The faculty member had primary responsibility for the educa-
tional content, process, and student evaluation. These roles became more fluid,
as instructors responded to students’ reflections and experiences. Both colleagues
began to contribute content, gave technical support, and engaged in discussion
through blogs and discussion forums. Students also became “experts” who posted
valuable resource materials on course topics. One student articulated this educa-
tional shift well in her final evaluation: “Previous coursework primarily involved
transmission of information by the professor to students. Assignments were sub-
51046_ch18_p181-186.qxd 5/19/11 2:23 PM Page 186
mitted to the instructor and evaluated. Other students in the class participated by
reading peers’ reflections and comments. Use of the network demanded engage-
ment of all class members with each other and with the instructors.”
Conclusion
Even with the specific issues experienced using Ning, the custom social network-
ing service successfully met all but one of the initial criteria set for a cloud-based
environment. Both projects involved multimedia files, requiring upload and stor-
age. Multiple leaders were involved with both projects and were able to work
jointly across locations successfully. Ning proved to be stable, without authenti-
cation error, and sufficiently low cost. The project leaders were freed to focus all
effort on project development with a short schedule and limited learning curve.
The service’s interface proved easy to use for project participants. No internal
campus resource had to be tapped or consulted. For all these reasons, there is a
great deal of potential in cloud-based web services like Ning where not only con-
tent needs to be delivered, but the space and functionality to foster community
and participant sharing are vital and where institutional resources for such efforts
are minimal. With menu-driven customization and domain name mapping (for
Plus and Pro level accounts), platforms such as Ning offer smaller educational or
cultural institutions a low-cost, feature-rich, easily implemented springboard for
service development.
186
51046_ch19_p187-190.qxd 5/19/11 2:24 PM Page 187
Chapter 19
Introduction
Libraries are increasingly looking to technology to solve issues of integrating instruc-
tion and reference services into academic courses, which may be offered partly or
wholly online. Cloud-based web conferencing software offers features such as chat,
screen sharing, and video to effectively present library instruction in an interactive, 187
engaging way. Many students may be unfamiliar with web conferencing, especially
older nontraditional students returning to the workforce or upgrading work skills. In
our search for the perfect web conferencing software, we thought we had found the
solution in a cloud application with all the features we were looking for in the free
version, only to find out a year later that the software had been sold and the service
was being discontinued for all users. Users with free accounts had three months’
notice before their accounts were closed, while paying customers were able to keep
their service until the end of their contract. As we reflect on our experiences, we find
that although some cloud-based applications offer low-cost, easy-to-use, feature-rich
software, they may not always be the best solution for critical library programs.
Background
The University of Washington (UW) is the largest university in Washington state,
with three campuses enrolling over 45,000 students. We serve the educational
needs of researchers, faculty, staff, students, and health professionals for the
WWAMI region (Washington, Alaska, Idaho, Montana, and Wyoming). Students
who use the Health Sciences Library are mainly upper-division students from all
over the state. Many of these students are older nontraditional students juggling
career, family, and education.
51046_ch19_p187-190.qxd 5/19/11 2:24 PM Page 188
The UW Health Sciences Library has the largest and most comprehensive col-
lection of health sciences materials in the Pacific Northwest. Our online web portal
offers off-campus access to a large number of health-related resources. We also
provide video and HTML tutorials on a variety of popular resources on our web-
site. In spite of the popularity of online, self-paced tutorials, there is often a need
for one-on-one training, especially for nontraditional students who may be return-
ing to the workforce or upgrading skills and who may not have the computer skills
of the younger generation. Many students, young and old, may also prefer to get
help from a librarian face-to-face, especially when working on a large research pro-
ject that requires accessing several different resources, all with different interfaces
and search methodologies. Finding the resources needed for research can be
daunting, involving hours of sifting through huge databases of information to find
a few key articles.
In our work in an academic medical library, librarians are called on more and
more to integrate library instruction with nontraditional courses such as evening,
online, or hybrid offerings. Communicating with students who we may see face-
to-face only once, if at all, is a huge challenge. When our nursing liaisons were
asked late in December to teach the research component in a large (50 students)
project-oriented winter course that would be largely offered online, we needed to
think creatively in order to come up with a distance education solution that would
offer the traditional benefit of in-person library instruction and consultation to off-
campus students. Web conferencing software offers the advantages of face-to-face
188 instruction, but at a distance. The solution we were looking for needed to be easy
to use and not be dependent on the fast network resources we are used to on
campus. We were looking for a solution that would allow us to do electronic chat
and video and screen sharing so that we could discuss the students’ research needs
and guide them through search strategies and resource discovery. All of this
needed to be done in an environment of diminishing budgets with limited techni-
cal support. The timeline also made it imperative that there be minimal training of
librarians needed. The software we were looking for also needed to be easy to
access for students of all ages and abilities with technology. We preferred applica-
tions with no software to be loaded on the student’s computer and limited setup
before use. The solution we came up with was to use the free version of a web
application called DimDim. This solution proved to be very successful, although
we encountered several challenges during and after implementation.
Selection Process
We originally chose a free cloud service because we had no budget for the project
and it needed to be implemented on very short notice. We had originally looked at
several web conferencing applications, including free cloud offerings, for-pay
cloud services, and commercial software. We chose DimDim because of the ease of
use for host and client, extensive features, and of course the free price tag. It also
51046_ch19_p187-190.qxd 5/19/11 2:24 PM Page 189
seemed to be well established and was built on open source software. DimDim
was advertised as “real-time collaboration made real easy.” DimDim allowed us to
e-mail a meeting invitation, which contained a clickable link that took the partici-
pant directly to the meeting. Screen sharing, document sharing, and a built-in
whiteboard were included. Participants could even share their computer screen
with the librarian, which is helpful when troubleshooting problems with accessing
resources. Meetings could be hosted and accessed online from anywhere at any-
time the librarian was available and were not dependent on specific hardware or a
particular browser.
After spending considerable time and effort choosing and implementing a
cloud application for online web conferencing for the nursing distance education
program, we were notified the software had been sold and we were given three
months to save our files before our login account would be closed on March 15,
2011. Those who paid for the added features of subscription service are sup-
ported only until their subscription runs out, and no new accounts are being
accepted. All users, free or paid subscription holders, were advised to download
files and conference transcripts from DimDim before the expiration period. Under
the license agreement we approved when signing up for a free DimDim account, it
was clear that they had the right to discontinue services at any time. We just had
not counted on that happening so soon.
Although it was built on open source software, the DimDim developers
decided to sell the code to a for-profit company after being offered several million
dollars. The code will be integrated into existing commercial software, expand- 189
ing an existing social platform for business communications to include web con-
ferencing. The new owners of the software, Salesforce.com, have announced that
besides shutting down the cloud version of the software, they will also not be sup-
porting the further development of the open source version, which has not been
well maintained recently. Although a fork of the code called MidMid was
announced (http://code.google.com/p/midmid), we are not aware of another ser-
vice using either the original DimDim or forked code. We do not have the staff to
further develop this software on our own. Salesforce.com has announced that with
the purchase of DimDim they plan to integrate the software into their existing
social collaboration platform, Chatter. This integration will allow them to offer an
integrated collaboration and communication platform in the future, but it is not
yet available.
After learning that DimDim was discontinuing their cloud services, we needed
to again form a committee to re-review web conferencing software. Fortunately,
due to a campuswide licensing program, we were able to purchase an Adobe Con-
nect account for a discounted rate. This software offers all of the features we were
looking for when we chose DimDim but is not as easy for librarians or students to
use. Unfortunately, we have to pay phone line charges to get a good audio connec-
tion, and sometimes students have trouble accessing the conference due to limita-
51046_ch19_p187-190.qxd 5/19/11 2:24 PM Page 190
Conclusion
The most important lesson we learned was that the best software for a particular
application is sometimes not free. We will probably not use a free cloud application
for a critical project again. We continue to use both free and inexpensive cloud
applications for online polling, sharing documents, IM, and surveys. Paying the
extra fee for a “pro” account with established cloud applications seems more
secure than relying on a completely free service. As we have seen with DimDim,
however, that is not always true. Another option would be to pay for cloud appli-
cations offered by established software companies, although again there is no guar-
antee that the specific product will be any more likely to be continued if it is not
deemed successful by the company.
While cloud applications can save a lot of money on computer hardware and
infrastructure as well as programming staff, there are some dangers to be aware of
when using it for critical library programs. Non-business critical programs are fine
with these risks, but for critical programs and websites, it may still be better to
host your own data or website and/or develop your own product using open
source software that is well established and has an active supporting community.
190 For more information about these types of issues, see the chapters by Rosalyn
Metz, H. Frank Cervone, and Carl Grant in Part I of this book.
Although we have purchased Adobe Connect web conferencing software and
are able to continue our program offering online reference assistance and screen
sharing, we still have problems with ease of use, audio lag time, and other quality-
of-sound issues. We have formed a committee to explore new web conferencing
options as they emerge. The field is very dynamic right now, and we have high
hopes that new technology will emerge to integrate online web conferencing into
the library workflow in a more seamless way and at an affordable price.
51046_ch20_p191-198.qxd 5/20/11 8:37 AM Page 191
Chapter 20
Introduction
Imagine for a moment a group of people gathered around a piece of artwork in a
gallery. The group may be collectively or individually discussing the piece, offering
opinions, background information, or interpretations of the exhibit. Each person
may have something of value to contribute, but louder opinions sometimes drown 191
out the more soft-spoken voices. Similarly, the experience of soliciting feedback from
students during instruction sessions can be disappointing when the same three stu-
dents (or worse, none at all) answer all the questions. What if there was a forum that
could offer equal time to everyone? The cloud technology tool VoiceThread offers
just that, and, as a bonus, it is easy to set up and fun to use. Offered as software as a
service (SaaS), VoiceThread allows users to upload still images, documents, or
videos to a custom display and invite others to view and comment on the materials.
These presentation modules, appropriately titled “VoiceThreads,” allow individuals
to share and exchange ideas around a common object. It is easy to integrate this
cloud technology into library instruction in a variety of ways.
Background
Keene State College in Keene, NH, is a public liberal arts college with an approx-
imate enrollment of 5,000 students and an active and growing information literacy
instruction program. At Mason Library, eight librarians teach more than 250 ses-
sions over the course of an academic year. This instruction takes place across all
majors and course levels. Because all incoming freshman are required to take the
interdisciplinary course Thinking and Writing, innovative approaches to reduc-
ing “IAKT” (I Already Know That) syndrome are always of keen interest.
51046_ch20_p191-198.qxd 5/19/11 2:24 PM Page 192
At Keene State College, the exploration and utilization of low-cost or free tools
in the classroom is a way to develop cutting-edge library instruction without being
financially burdened or committed. The online tool VoiceThread allows collabo-
rative conversations around multimedia presentations, encouraging participation
from students in nontraditional ways. The discussions take place in the cloud
without the need for downloading or hosting software, something to consider
when the library does not have access to a server. Despite the relative simplicity of
the program application, VoiceThread can be used in a variety of ways depending
on the desired learning outcomes of the instructor using it. The flexibility and
adaptability of the program allows for easy customization and application beyond
the four walls of the traditional classroom. Promoting an ongoing collaborative
educational experience rather than a static resource page, VoiceThread can be
embedded in course management software, accessed through the web, or linked
from subject guides. This functionality provides the opportunity to enhance con-
tent and build dialogue around course materials and subjects extending beyond
the class session and even the campus itself.
Incorporating cloud technologies, like VoiceThread, into instruction helps
engage beginning students and focus more advanced library users on the session
content. It also gives librarians and professors the ability to assess student learning
outcomes and teaching effectiveness.
Selection Process
192 Although there are other programs similar in focus to VoiceThread, and while
VoiceThread itself is a mashup of several Web 2.0 concepts (online communi-
ties/forums, Facebook, Slideshare, YouTube, etc.), it is still unique in how it builds
upon these existing frameworks. We chose VoiceThread because of the low cost
and the way modules can be created using technologies we are already familiar
with, including PowerPoint, Jing, and LibGuides. VoiceThread has been available
since 2007, winning awards for its innovative application of collaborative software.
It has also been featured periodically on educational blogs (http://www
.webware.com/html/ww/100/2008/winners.html and http://sloanconsortium.org/
effective_practices/voicethread-enhanced-community-increased-social-presence-
and-improved-visual-lea).
We see students for a relatively limited time during the semester, and
VoiceThread provides a way to encourage participation in the class community.
This tool is valuable because of the distinct way it engages students during class
and continues the conversation after the class session and beyond the physical
space of the library in the computing cloud. However, before using this, or any
other technology, it is important to consider the learning outcomes for which
the tool would be useful. VoiceThread isn’t used for every class, only those in
which the features of the tool enhance or help to achieve student learning out-
comes.
51046_ch20_p191-198.qxd 5/19/11 2:24 PM Page 193
We chose to use the free account of VoiceThread, which allows each registered
account up to three VoiceThreads. This means that a classroom can only have
three “sessions” in play at any time. This may be a problem if librarians anticipate
using this tool in multiple classes. Like many cloud programs, VoiceThread
requires payment to unlock greater functionality or storage space.
As much as we like using VoiceThread, we have not yet chosen to upgrade to a
paid account due to the cost of doing so. For Higher Education users, a single
license for a Pro account (including the ability to manage up to 50 Basic accounts)
is $99 for one year, while a departmental account (up to 10 Pro accounts and 250
Basic accounts) currently costs $699 per year. Site-wide licenses are available upon
request and vary on the size of the institution. With the increase in costs come
greater functionality, storage space, and user account management features not
available at the lower tiers. Additionally, there are account options available for
users not affiliated with Higher Education.
After creating a VoiceThread account, it is easy to create modules. Materials
are uploaded to VoiceThread from one’s computer, media sources (Facebook,
Flickr, the New York Public Library, or another VoiceThread), a URL, or an
attached webcam. Once items are uploaded, VoiceThread can be shared with
others via a custom URL, or it can be embedded in a variety of social media plat-
forms using the code automatically generated. Publishing options allow a user to
choose privacy settings. Playback options are somewhat adjustable for the free
account, with greater functionality available to the paid account users. For a fee,
archival quality versions of the VoiceThread can be exported. 193
Implementation
Jennifer’s Implementation: VoiceThread to Encourage
Community Growth as Part of an Instruction Toolkit
After providing one-shot library instruction sessions in connection with program-
matic and student learning outcomes specific to that course, students still have ques-
tions and need more help. I use VoiceThread to build course support modules for
students to revisit the materials presented in class. These modules reinforce informa-
tion literacy skills to help students gain mastery of the material. They also give help
specific to library technology tools such as databases and the catalog. The modules
include topics such as how to search the catalog, how to authenticate access to data-
bases from off campus, and how to search databases efficiently and effectively. I
uploaded a screenshot tutorial on these topics into VoiceThread and offered the
embedded VoiceThread for posting in the course management system for that class.
Students were encouraged via e-mail to discuss and ask questions about the concepts
using the VoiceThread module with voice, text, or video. Contributing students
developed a community beyond the physical face-to-face interactions of the class-
room and extended learning to the virtual world, sharing their own perspectives.
51046_ch20_p191-198.qxd 5/19/11 2:24 PM Page 194
This exercise works well to engage students and encourage peer-to-peer eval-
uation of the class topics, but I wondered if participation could be further
enhanced through the use of cloud technology. The added bonus of using such a
tool would be the enduring record of the collective work of the class (at least for
the short term).
Having worked with VoiceThread previously, I thought it offered the perfect
opportunity to boost the traditional tried-and-true approach to the virtual level.
This program could move the discussion to the online environment, which might
encourage more students to participate and encourage higher quality participa-
tion. I imagined that each student would be able to get feedback from others in
the class through VoiceThread’s text commenting capabilities. With this in mind,
I adapted the previous exercise as follows:
1. Either ahead of time or during the first part of class (though beforehand
saves time), each student e-mails the librarian a PowerPoint slide with
two pieces of information: the topic statement/question and why this
topic is of individual interest.
2. The librarian combines the slides into one PowerPoint file and uploads
it to VoiceThread.
3. A brief tutorial is provided to the class on how to use VoiceThread,
with the librarian offering a live example of how to add a text comment
to a slide.
4. Students are encouraged to supply at least one comment/item for con- 195
sideration/question on each person’s slide, with the librarian also par-
ticipating.
5. After a set amount of time, students find their own slides and view what
others have suggested. The suggestions are condensed into keywords
that are employed in new search strategies in the library’s online data-
bases.
Preparation before the session included creating a generic VoiceThread
account using a free e-mail account I set up for this purpose. Every student was
able to successfully log in to the software using the same account. However, the
unfortunate drawback to this approach was that each comment came from “the
same” user; it is visually more appealing while using VoiceThread to see the indi-
vidual icon of each user around the periphery of the image in the center. At the
conclusion of the session, I was able to quickly export the code for the
VoiceThread community we had created together and embed it into the class
LibGuide, showing the students where to find it before the end of the session.
success. VoiceThread successfully achieved the goal of using a low-cost, easy setup
cloud technology to encourage participation and collaboration in library instruction.
Jennifer’s implementation of VoiceThread supports library instruction ses-
sions by encouraging nontraditional participation. Students less likely to speak up
in class or ask questions were given the opportunity to do so in a low-risk envi-
ronment. Creating and embedding presentation modules is straightforward, allow-
ing quick setup and ease of use. Although higher participation rates were hoped
for, there are some reasons for low usage. The necessity to create an account to
make comments and ask questions may have discouraged students. Although the
class professor was interested and supportive, VoiceThread might not have been
promoted well during the class. In the future, students will be given time in class to
create an account so they can try out the system and see how easy it is. This will
also allow the professor a hands-on chance to become involved during class. Pro-
fessors will also be asked what types of VoiceThread presentation modules are
needed for the class. This will give the professor buy-in from the start, increasing
the marketing of the materials for use.
Kara was pleasantly surprised in doing the analysis of her implementation to
see that each student had commented on an average of about 11 slides per class
and that the feedback provided was meaningful. Thought-provoking concepts for
further study and excellent critical questions were the norm in peer-to-peer com-
ments. This showed that VoiceThread did increase student participation and col-
laboration in library instruction.
196 Students were completely engaged in the activity and the instruction class-
room was abuzz with the flutter of keyboard activity. Informal feedback indicated
that the exercise did help students think about aspects of their topics they had not
previously considered. In observing search behavior after the VoiceThread exer-
cise, some students were expanding their keyword usage beyond the few they had
been using before the session. To maximize the session and take advantage of the
community feel of VoiceThread, it will be beneficial in the future to set up a
VoiceThread account for each individual student. Additionally, it would save time
to gather student topics and questions about topics before the day instruction is to
take place. This gives the librarian time to build the class PowerPoint ahead of
time, including a final slide soliciting feedback on the exercise.
Conclusion
Using the cloud technology tool VoiceThread has been easy and fun. We look for-
ward to developing other projects using this tool to enhance and support library
instruction in the future. Here are some recommendations we have developed for
the use of VoiceThread:
• Discussing and recommending resources: Take a screenshot of a book
cover or take a picture of someone holding an item. Record a brief
51046_ch20_p191-198.qxd 5/19/11 2:24 PM Page 197
review of why the source is a good choice for a given research topic. Stu-
dents then find a resource of their own to recommend to the group and
record a message describing their choice.
• Class discussion: Post slides from library instruction sessions with
added questions. Have students record their answers. This keeps a dis-
cussion going beyond class, allowing nonverbal processors to partici-
pate.
• Library orientation: Upload pictures of the library. Have students view
this virtual tour and ask questions or make comments.
• Tutorials: Capture screen images to provide a step-by-step tutorial
incorporating the visual with the auditory on how to find and use library
resources.
• Blended learning: Use this platform to engage a group of people who
have never met before. The pictures and voice recordings allow unique
discussion opportunities for virtual and blended classrooms.
• Student learning outcome assessment: Upload a two-picture
slideshow asking students these questions, “What did you learn today?”
and “What questions do you still have?” to create ongoing or end-of-
session evaluations.
Overall, integrating VoiceThread into our instruction sessions has been a suc-
cess. It facilitates collaborative teaching and learning among the students, profes-
sors, and librarians by supporting conversations in multiple modes around a 197
common object. It encourages participation within the classroom and allows learn-
ing to take place beyond the physical classroom. VoiceThread also gives students
control over their learning experience as they revisit presentation modules. We
would use free or low-cost cloud technologies such as VoiceThread again because
they give libraries the chance to experiment with new technologies without being
financially invested. Cloud technologies are more convenient to use than tradi-
tional software packages because they can be learned and utilized in a short period
of time. The time-consuming work of setting up or configuring a server, or learn-
ing a fully featured program is not an issue with the proliferation of such low-cost
and low-barrier programs hosted on the web. We have used VoiceThread to
enhance instruction sessions and look forward to using it in the future for teaching
information literacy skills and addressing student learning outcomes.
51046_ch20_p191-198.qxd 5/19/11 2:24 PM Page 198
51046_ch21_BM_p199-204.qxd 5/20/11 8:38 AM Page 199
Marshall Breeding is the Director for Innovative Technology and Research at Vanderbilt
University. He is the creator of Library Technology Guides (http://www.librarytechnol-
ogy.org). In 2010 he authored Next-Gen Library Catalogs for Neal-Schuman and has writ-
ten or edited five other books. Breeding writes a monthly column for Computers in
Libraries, is a Contributing Editor for Smart Libraries Newsletter, has authored Library
Journal’s “Automation System Marketplace” since 2002, and has written many other arti-
cles and book chapters.
H. Frank Cervone is the Vice Chancellor for Information Services at Purdue University
Calumet, Hammond, Indiana (http://www.cervone.com). The author of four books and a
frequent speaker at library and information technology conferences, he writes a recurring
column in OCLC Systems and Services. He is the LITA Representative to the International
Federation of Library Associations and Agencies. Cervone holds a Master of Science in
Education with a specialization in Online Teaching and Learning from California State
University as well as a PhD in Management and Information Systems from Northcentral
University.
Karen A. Coombs (http://www.librarywebchic.net) is the Product Manager for the OCLC
Developer Network. Coombs has presented at many conferences on topics related to web
services, mashups, and open source software in libraries. She is the author of two books—
Library Blogging with Jason Griffey and Open Source Web Applications in Libraries with
Amanda Hollister. Coombs earned her Master of Library Science and Master of Science in
Information Management from Syracuse University, Syracuse, New York.
John Davison is Assistant Director for Digital Resource Management at the Ohio Library
200 & Information Network (http://www.ohiolink.edu/). Davison’s research interests include
cloud-based distributed computing, open source software development, and digital object
preservation. Davison earned a Master of Library and Information Science from the Kent
State School of Library and Information Science and a Bachelor of Arts from the Ohio
State University.
Leland R. Deeds is Librarian for Academic Computing Support at Union Presbyterian
Seminary, Richmond, Virginia. Deeds has presented on topics related to web technolo-
gies, assessment, and training. Deeds earned a Bachelor of Arts in Religion from Bard Col-
lege, a Master of Theological Studies from Emory University, and a Master of Science in
Information Systems from the University of Tennessee, Knoxville.
Jennifer Diffin (jdiffin@umuc.edu) is the Assistant Director for Systems and Access Ser-
vices at the University of Maryland University College in Largo, Maryland. Diffin has writ-
ten articles and presented at national and regional conferences on several topics, including
the effective use of Microsoft SharePoint in libraries. Diffin earned a Master of Library
Science from Simmons College, Boston, and a Bachelor of Science in computer science
and math also at Simmons College.
Jennifer Ditkoff is the Collection Development Librarian at Keene State College in Keene,
New Hampshire. She has presented at several conferences on using low-cost, low-barrier
technologies in the classroom. She earned her Master of Science in Library and Informa-
tion Science from Syracuse University in Syracuse, New York.
51046_ch21_BM_p199-204.qxd 5/19/11 2:24 PM Page 201
Heidi M. Nickisch Duggan is the Associate Director at the Galter Health Sciences Library
at Northwestern University’s Feinberg School of Medicine. Nickisch Duggan holds a Bach-
elor of Arts from the University of Nebraska–Lincoln, a Master of Arts in Library Science
from the University of Missouri–Columbia, and a Master of Science in Communication
from Northwestern University. Her research interests include library leadership, organiza-
tional change management, and scholarly communication.
Nicole C. Engard (nengard@gmail.com) is the Director of Open Source Education at
ByWater Solutions. In addition to her daily responsibilities, Engard keeps the library com-
munity up-to-date on web technologies via her website “What I Learned Today. . . .”
(http://www.web2learning.net). In 2009 Engard was the editor of Library Mashups, a book
published by Information Today, and in 2010 she authored Practical Open Source Soft-
ware for Libraries, a book published by Chandos.
Michelle Frisque is Head of Information Systems, Galter Health Sciences Library, North-
western University's Feinberg School of Medicine. Frisque is a former President of the
Library & Information Technology Association (LITA), a division of the American Library
Association. Frisque graduated with a Master of Library and Information Science from
the University of Pittsburgh and is currently pursuing a Master of Learning and Organiza-
tion Change at Northwestern University. Frisque has written articles and presented at con-
ferences on library IT management issues.
Ann Whitney Gleason is the Head, Systems at the University of Washington Health Sci-
ences Library. Her research interests currently center on the use of educational technol-
ogy in libraries. Gleason worked as an educational technology specialist for several years
before receiving her Master of Library and Information Science from the University of 201
Rhode Island. She received a Bachelor of Arts in Education from The Evergreen State
College.
Carl Grant is Chief Librarian at Ex Libris. Grant’s commitment to libraries and librarian-
ship is known from his participation in ALA; ACRL, LITA, Exhibits Round Table; and on
the board of the NISO, where he has served as board member, treasurer, and chair. Library
Journal has also named Grant an industry notable. Grant holds a Master’s degree in infor-
mation and library science from the University of Missouri at Columbia.
Edward Iglesias is the Systems Librarian at Central Connecticut State University and part
of the ERIS (Electronic Resources and Information Systems) group in a variety of capaci-
ties, ranging from ILS Systems Administration and Coordination to overseeing digital ini-
tiatives and encouraging technical curiosity. He just authored a book, An Overview of the
Changing Role of the Systems Librarian: Systemic Shifts, published by Chandos. Iglesias
earned a Master of Library and Information Science at the University of Texas and a
Master of Arts in English from Texas A&M International University.
Cindy Kissel-Ito is Director of the Extended Campus Program and Affiliate Assistant
Professor of Christian Education at Union Presbyterian Seminary, Richmond, Virginia.
Kissel-Ito’s academic interests focus on supporting student investigation of the curricu-
lum and developmental theory that shape religious education theory and practice. Kissel-
Ito received a PhD from Union Theological Seminary–Presbyterian School of Christian
51046_ch21_BM_p199-204.qxd 5/19/11 2:24 PM Page 202
Education, a Master of Divinity from Wesley Theological Seminary, and a Bachelor of Arts
in Anthropology from Albany State University.
Ann Thomas Knox is Director of the Instructional Resource Center at Union Presbyter-
ian Seminary, Richmond, Virginia. Knox has a particular interest in ways that web tech-
nologies can be used in teaching. Knox earned a Bachelor of Arts in History from Austin
College, Sherman, Texas, and a Master of Arts in Christian Education from the Presbyter-
ian School of Christian Education, Richmond.
Anne Leonard is an Instruction & Reference Librarian and the Coordinator of Resource
Sharing at New York City College of Technology of the City University of New York. Her
varied professional interests include the faculty and professional status of academic librar-
ians and academic integrity among undergraduates. She earned a Master of Library Sci-
ence from the University of Texas at Austin and a Master of Science in Urban Affairs from
Hunter College.
Rosalyn Metz (http://www.rosalynmetz.com) is a project manager for Viget, a web con-
sulting firm located near Washington, DC. She has written about cloud computing on her
blog (http://www.rosalynmetz.com/ideas) and for EDUCAUSE Quarterly. Additionally,
Metz has presented at multiple conferences on the cloud, including ALA Annual. She has
a Master of Library Science from the University of North Carolina at Chapel Hill and a
Bachelor in Political Science from George Washington University.
Robin Elizabeth Miller (millerob@uwec.edu) is Reference and Instruction/Government
Publications Librarian at the University of Wisconsin–Eau Claire. Miller earned a Master
of Science in Library and Information Science from the University of Illinois at
202 Urbana–Champaign and a Bachelor of Arts in Political Science from the University of
Oregon.
Erik Mitchell (http://erikmitchell.info) is the Assistant Director for Technology Services at
the Z. Smith Reynolds Library at Wake Forest University. He holds a Master of Library
and Information Science from the University of South Carolina and a PhD in Information
and Library Science from the University of North Carolina at Chapel Hill. His research
focus includes metadata and cloud computing and the role that these technologies play in
information experiences. His most recent work focused on exploring the role of cloud
computing in library IT. He currently writes a recurring column for the Journal of Web
Librarianship.
Dennis Nangle (dnangle@umuc.edu) is a Document Management Technician at Univer-
sity of Maryland University College in Largo, Maryland. Nangle has written multiple arti-
cles and presented at local and national conferences on the topic of establishing effective
library workspaces using Microsoft SharePoint. Nangle also worked on a project through
ALA’s Emerging Leaders program, which provided best practices for LITA members
working in ALA Connect. Nangle earned his Master of Library Science at the University of
Maryland, College Park.
Christopher R. Nighswonger is Network and Systems Director at Foundations Bible Col-
lege and Theological Seminary, Dunn, North Carolina (http://www.foundations.edu/).
Nighswonger has been involved in a variety of segments of the IT field for 25 years and in
51046_ch21_BM_p199-204.qxd 5/19/11 2:24 PM Page 203
active development for the Koha open source ILS project since November 2007. In Feb-
ruary 2010, he was elected Release Maintainer for Koha 3.2. Nighswonger earned his
Master of Theology from Foundations Bible College and Theological Seminary, Dunn.
Karen A. Reiman-Sendi (karsendi@umich.edu) is the Digital Information Services
Librarian at the University of Michigan Library. Reiman-Sendi has presented and pub-
lished on topics such as Library 2.0 applications, web projects, information literacy, virtual
reference services, collection assessment, and integrating desktop applications for public
services desks. Reiman-Sendi earned a Master of Information and Library Science from
the University of Michigan and a Bachelor of Arts in German Languages and Literature
from the University of Toledo.
Christinger R. Tomer, University of Pittsburgh Associate Professor, teaches asynchro-
nous and blended classes. He holds degrees from the College of Wooster and Case West-
ern Reserve University. Dr. Tomer’s areas of specialization include open source computing
and scientometrics.
Kenneth J. Varnum (varnum@umich.edu) is Web Systems Manager at the University of
Michigan Library. Varnum has presented at national conferences on topics including
Drupal, user-generated content, personalized search, and website redesign. Varnum earned
Masters of Information and Library Science and Russian Studies from the University of
Michigan and his Bachelor of Arts from Grinnell College.
Kara Young is the Systems Librarian at Keene State College in Keene, New Hampshire.
Young enjoys working with technology every day and sharing her experiences with low-
cost, low-barrier web applications by presenting at workshops and conferences. She earned
her Master of Science in Information Systems and Master of Public Administration 203
degrees from the State University of New York at Albany.
51046_ch21_BM_p199-204.qxd 5/19/11 2:24 PM Page 204
51046_ch22_IN_p205-214.qxd 5/19/11 9:47 AM Page 205
Index
Page numbers followed by “f ” indicate figures; those followed by “t” indicate tables.
7 Signposts to Library 2.0 (custom social Simple Storage Solution (S3), 13, 22,
network), 184 146, 149, 150t
21st Century Cutting-Edge Technology Virtual Private Cloud (VPC), 24, 26
Service for 2011 (award), 111 American Psychological Association
360 Link (link resolver from Serials (APA) citations format, 94, 102f
Solutions), 21 Analytics, 38, 49, 57
360 Search (federated search product Android (operating system), 103, 104,
from Serials Solutions), 75 105
Android Market, 16
A Announcements and SharePoint, 137
Adobe Connect (web conferencing APA citations format. See American
software), 189, 190 Psychological Association (APA)
Aggregated indexes, 51 citations format
building, 82–85 APIs, 4, 14, 15, 16, 19, 40, 120 205
and discovery systems, 77–78, 81–82 from Amazon, 39, 112
vs. federated search, 75–76 definition, 39
Alma (next-generation library services for real-time data for discovery
framework from Ex Libris), 50 interfaces, 78
Amazon (firm), 13, 39 WorldCat Search API, 94–96, 98, 102,
Amazon Machine Images (AMIs), 23, 111 103, 104, 105, 106
and the DRC software, 114–115, 115f See also WorldCat Registry API
See also Virtual machines (VMs) Application programming interfaces
Amazon Web Services (AWS), 15, 17f, 23, (APIs). See APIs
38, 80 AquaBrowser Library (discovery service
and DSpace, 109 from Medialab Solutions/Serials
Elastic Compute Cloud (EC2) (virtual Solutions), 79–80
machine provisioning system), 22, Archives of the History of American
43, 50, 51, 112 Psychology, The, 110
as fast solution, xii, 13, Archivists’ Toolkit (digital
registration, 112–113, 113f archiving/repository system), 66t,
with VMWare’s vCloud service, 26, 26f 148
Management Console, 18, 19f, Archon (digital archiving/repository
112–117, 113f system), 64, 66t
Relational Database Service (RDS), Australia, 7
13, 18 Azure (hosting from Microsoft), 22
51046_ch22_IN_p205-214.qxd 5/19/11 9:47 AM Page 206
B Cloud computing
Backups, 8, 22, 126–127, 143, 145 definition of, 3, 14, 20, 23, 31
BagIt (transmission guidelines from future, xii, 33, 56–57, 59, 61
Library of Congress, California as green, 49
Digital Library, and Stanford vagueness of definitions, 30
University), 148 Cloud Computing Risk Assessment
Bandwidth, 5, 32, 54, 60 (ENISA), 24
Bar codes, 98, 104, 105 Cloud Security Alliance. See CSA (Cloud
BiblioCommons (discovery service), 79 Security Alliance)
Big Switch, The (N. Carr), 16 Collaboration, 49, 62–63
Blacklight (open source discovery service College & University Disability Access
from the University of Virginia), 80 Collaborative (Ohio), 110
Blogs and SharePoint, 135, 139 Colorado School of Mines Library, 102
BookMinder prototype (mobile Communication as a service, 23
application), 103, 104 Community cloud (deployment model),
Box.net (file sharing software), 128–129 23, 24–25, 134
Brandeis University Library, 103 Compliance, 123–124, 126, 128, 131
Branding, 53, 54, 83, 110 Comprehensive indexes. See Aggregated
Bryn Mawr Library, 102 indexes
Burritt Library, 144 Computer laboratories (LIS education),
62
C Computing, history, xi, 37–38, 47
C3. See Cumulus Cloud Computing (C3) Content management systems, 64, 65, 65t
Calendar software, 136. See also Google CONTENTdm (digital collection
206 Calendar (calendar software) management software from OCLC),
Cambridge Information Group, 80 63, 65, 66t, 144
CampusBooks (mobile application), 103, Contracts for cloud services, 5, 6–7, 8, 9,
105, 105f 91, 187. See also SLAs
Canada, 6, 7 Coombs, Karen A., 101, 103
Carr, Nicholas, 16, 39 Core Access License (CAL) (from
Center for Applied Research (ECAR) Microsoft), 133
(EDUCAUSE), 39 Corporate libraries, 56
Central Connecticut State University, 144 Cost savings, 5, 34, 40, 48, 49, 60
Central Search (federated search product and Koha, 89
from Serials Solutions). See 360 and S3, 151
Search (federated search product Council on Library and Information
from Serials Solutions) Resources, 78
Chatter (social collaboration platform), CSA (Cloud Security Alliance), 52–53
189 Cumulus Cloud Computing (C3), 25
Chicago citations format, 94, 102f CUNY system, 166
Chrome (web browser), 130 CWIS (digital archiving/repository
Citation-only searching, 83 system), 66t
Cite This (tool from OCLC), 101–102, 102f
City Tech. See New York City College of D
Technology (City Tech) DaaS (data as a service), 4, 49, 53, 93
Clarke, Arthur C., 47, 57 Darien Library, 80
51046_ch22_IN_p205-214.qxd 5/19/11 9:47 AM Page 207
Index
eXtensible Catalog (open source discovery for productivity, 38, 128, 168
system from the University of See also Google Forms (spreadsheet and
Rochester), 80 form software)
Google Documents. See Google Docs
F (file-creation and file-sharing
Facebook (social networking site), 3, 38, software)
102, 162, 192, 193 Google Forms (spreadsheet and form
Faceted navigation, 77 software), 173, 174, 175–179, 177f
Family Educational Rights and Privacy Google Image Search, 15
Act. See FERPA (Family Educational Google Maps, 103
Rights and Privacy Act) Google Scholar, 83
FAST (Faceted Application of Subject Google Spreadsheets, 38, 177
Terminology), 96 Google Storage for Developers, 13, 23
Federated search, 74–76, 110 Gorman, Michael, 61
Fedora (digital archiving/repository Greenstone (digital archiving/repository
system), 66t, 120 system), 66t
Fedora Commons, 64, 119 GroupWise (e-mail and calendar system),
FERPA (Family Educational Rights and 165, 167–168
Privacy Act), 6, 52, 126 GSA Form and Genre Terms, 96
File sharing, 123–124, 124–126, 137, 159
Files as a service, 23 H
Files in This Item (blog), 121 Harvard citations format, 94, 102f
Firefox (web browser), 130 HathiTrust (digital library), 95
Flexibility, 4, 60, 63 Health Insurance Portability and
208 Flickr (photo-sharing site), 16, 184, 185, Accountability Act. See HIPAA
193 (Health Insurance Portability and
Florida Digital Archive, 143 Accountability Act)
Florida State University, 63 Heroku (hosting), 22
Free services, 5, 33, 187, 190 Hewlett Packard (HP), 145
Full-text searching, 83 HIPAA (Health Insurance Portability and
Accountability Act), 6, 126
G Horowhenua Library Trust, 66t, 87–88
Gartner group, 34, 47, 50 Houston, Drew, 160
getHistory request type, 95 Hybrid cloud (deployment model), 23,
Gmail (e-mail), 3, 165 25–26, 50
Go Daddy (web hosting), 22
Google (search engine), 85 I
Google, Inc., 13, 23 IaaS (infrastructure as a service), 20,
Google App Engine, 13, 22, 50 22–23, 41, 50
Google Apps for Education, 40, 171 and library systems, 40, 41, 41t, 42t, 89
Google Apps suite, 129, 171 IAKT (I already know that) syndrome,
Google Calendar (calendar software), 165 191
for group scheduling, 168–170, 169f, IBM (computing vendor) as private cloud
171 provider, 25
Google Docs (file-creation and file-sharing iBookshelf (mobile application), 103, 104,
software), 5, 13, 16, 130. 104f
51046_ch22_IN_p205-214.qxd 5/19/11 12:12 PM Page 209
Index
Index
Online Public Access Catalogs (OPACs). Pew Internet Trust, 37, 39, 40
See Online catalogs Photo sharing, 138. See also Flickr
OPACs. See Online catalogs (photo-sharing site)
Open Archives Initiative (OAI), 76 Pic2Shop (mobile application), 103,
Open Archives Initiative—Protocol for 105–106, 106f
Metadata Harvesting (OAI-PMH). Ping.fm, 162
See OAI-PMH (protocol) PIPEDA. See Personal Information
Open Cloud Manifesto, 55 Protection and Electronic
Open Conference Systems (Public Documents Act (PIPEDA)
Knowledge Project software), PKP Harvester. See Open Harvester
67t Systems (Public Knowledge Project
Open Content Alliance, 95 software)
Open Harvester Systems (Public Platform as a service (PaaS). See PaaS
Knowledge Project software), (platform as a service)
67t Platform neutrality, 48
Open Journal Systems (Public Knowledge Plone, 65t
Project software), 67t Portico (e-journal archive), 143
OpenBiblio (open source ILS), 65, 66t PowerPoint (presentation software), 192,
OpenSearch (protocol for sharing search 195
results), 94 PREMIS (data dictionary from the Library
OpenURL (link resolvers), 74, 79, 96, of Congress), 148
97–98, 99 Primo (discovery service from Ex Libris),
Oracle (computing vendor) as private 79
cloud provider, 25 Primo Central (web-scale discovery
OS X (operating system from Apple), 21, service from Ex Libris), 81–82 211
175 Privacy in the cloud, 5–7, 51, 52, 60, 182,
OSHEAN, the Rhode Island Research 193
and Education Network, 24–25 and Dropbox, 161, 163, 164
Outlook (e-mail/calendar software from and FERPA, 6, 126
Microsoft), 137 and Google Calendar, 168, 170
Outsourcing of cloud services, 8, 17, 24, Private cloud (deployment model), 23, 25,
30, 91 50, 133–134
Provisioning of services, 29
P Public cloud (deployment model), 23–24,
PaaS (platform as a service), 20, 21–22, 50, 134
40, 41t, 42t, 50 as platform for Koha (open source ILS),
Patriot Act, 7, 52, 56 89–90
Pazpar2 (metasearching middleware web Public Knowledge Project, 65–66
service from Index Data), 75 Public Knowledge Project Open Harvester
Pearson partnership with Ning, 185 Systems. See Open Harvester
Peer Reviewed Journal script, 101, 101f, Systems (Public Knowledge Project
107 software)
Personal Information Protection and
Electronic Documents Act Q
(PIPEDA), 6 Qualtrics (survey software), 174
Pew Internet & American Life, 15 Quotations Page, 47, 57
51046_ch22_IN_p205-214.qxd 5/19/11 9:47 AM Page 212
Index